CN114723613A - Image processing method and device, electronic device, storage medium - Google Patents

Abstract

The disclosure relates to an image processing method and device, an electronic device and a storage medium. Wherein, the method comprises the following steps: acquiring a to-be-processed image of a brightness-chrominance space domain, and performing down-sampling operation on the to-be-processed image to obtain a low-frequency image of the to-be-processed image; respectively determining the chromaticity weight of each first pixel according to the brightness difference degree of each first pixel and each first neighborhood pixel in the image to be processed, wherein the chromaticity weight of each first pixel in each first pixel is positively correlated with the corresponding brightness difference degree; and performing upsampling operation on the low-frequency image based on the chromaticity weight corresponding to each first pixel, and synthesizing the image obtained through the upsampling operation with the image to be processed to obtain a processed image. By the method, the chromatic value of the image can be adjusted in the image processing process, and the condition that color edge color overflows in the processed image is avoided.

Description

Image processing method and device, electronic device, storage medium

technical field

The present disclosure relates to the field of image processing, and in particular, to an image processing method and device, an electronic device, and a storage medium.

Background technique

In the process of collecting, transmitting, and processing images, electronic devices are inevitably disturbed by various factors, resulting in the acquired images accompanied by noise signals. The noise signal accompanying the image mainly includes two kinds of noise signal, namely luminance noise and color noise (also known as: chrominance noise).

In the related art, a relatively mature noise reduction technology has been developed for the processing of luminance noise. However, for the color noise in the image, it is still difficult to perform effective noise reduction processing, especially at the color edge with a large color span, color overflow or blurring often occurs.

SUMMARY OF THE INVENTION

In view of this, the present disclosure provides an image processing method and device, an electronic device, and a storage medium. In the process of upsampling a low-frequency image, the chromaticity value of the low-frequency image can be processed according to the brightness distribution of the high-frequency image. Adjustment to avoid color bleed at the edges of the upsampled image.

To achieve the above object, the present disclosure provides the following technical solutions:

According to a first aspect of the present disclosure, an image processing method is proposed, including:

Acquiring a to-be-processed image in the luminance-chrominance space domain, and performing a downsampling operation on the to-be-processed image to obtain a low-frequency image of the to-be-processed image;

The chromaticity weight of each first pixel is determined according to the degree of difference in luminance between each first pixel in the image to be processed and the respective first neighborhood pixels, wherein each of the first pixels, each The chrominance weight of the first pixel is positively correlated with the corresponding luminance difference;

An up-sampling operation is performed on the low-frequency image based on the chrominance weight corresponding to each first pixel, and the image obtained through the up-sampling operation is synthesized with the to-be-processed image to obtain a processed image.

According to a second aspect of the present disclosure, an image processing apparatus is proposed, comprising:

an acquisition unit that acquires an image to be processed in the luminance-chrominance space domain, and performs a downsampling operation on the image to be processed to obtain a low-frequency image of the image to be processed;

The determining unit determines the chrominance weight of each first pixel according to the brightness difference between each first pixel in the to-be-processed image and the respective first neighborhood pixels, wherein, in each first pixel , the chrominance weight of each first pixel is positively correlated with the corresponding luminance difference;

The synthesis unit performs an up-sampling operation on the low-frequency image based on the chrominance weight corresponding to each first pixel, and synthesizes the image obtained through the up-sampling operation with the to-be-processed image to obtain a processed image.

According to a third aspect of the present disclosure, there is provided an electronic device, comprising:

processor;

memory for storing processor-executable instructions;

Wherein, the processor implements the method according to the first aspect by executing the executable instructions.

According to a fourth aspect of the present disclosure, there is provided a computer-readable storage medium having computer instructions stored thereon, the instructions, when executed by a processor, implement the steps of the method according to the first aspect.

In the technical solution of the present disclosure, an image processing method is provided. On the one hand, the method performs a downsampling operation on the obtained image to be processed in the luminance-chrominance space domain to obtain a corresponding low-frequency image; The luminance difference degree of the pixels is assigned a chrominance weight to each of the included first pixels, and the chrominance weight of each first pixel is positively correlated with the corresponding luminance difference degree. On this basis, an up-sampling operation can be performed on the low-frequency image based on the chrominance weight corresponding to each first pixel in the image to be processed to obtain an image with the same size as the image to be processed, and the image and the image to be processed are merged , to get the processed image.

It can be understood that in the image field, the change of the color in the image will affect the change of the brightness, and the change of the color is larger, and the change of the brightness is usually larger. In other words, changes in brightness in an image can reflect changes in color. For any pixel, if the degree of difference in luminance with neighboring pixels is greater, it means that the difference in chrominance between any pixel and its neighboring pixels is generally greater. In the present disclosure, the chrominance weight of any pixel is positively correlated with the corresponding degree of luminance difference, which is equivalent to assigning a higher chrominance weight to a pixel in an area with large luminance variation (at the color edge). On this basis, up-sampling the low-frequency image by the determined chrominance weight is equivalent to increasing the chroma value of the pixel at the color edge in the low-frequency image, making the color change at the color edge more prominent, thereby avoiding The problem of color overflow at the color edge in the related art.

In short, the present disclosure identifies the color edge in the image by analyzing the brightness change in the image, and then assigns a higher chrominance weight to the pixel at the color edge, so that the pixel at the color edge The difference between the chromaticity value of the pixel and the chromaticity value of the pixel where the color is flat increases, which highlights the difference between the color edge and the color flat, and solves the problem of color overflow at the color edge in the related art.

Description of drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the disclosure and together with the description serve to explain the principles of the disclosure.

1 is a flowchart of an image processing method according to an exemplary embodiment of the present disclosure;

FIG. 2 is a schematic diagram of an image to be processed according to an exemplary embodiment of the present disclosure;

FIG. 3 is a schematic diagram of a low-frequency image according to an exemplary embodiment of the present disclosure;

4A is a schematic diagram of an image obtained by an upsampling operation according to an exemplary embodiment of the present disclosure;

4B is a schematic diagram illustrating a comparison between an upsampling operation performed by bilinear interpolation and a joint guided upsampling operation according to an exemplary embodiment of the present disclosure;

FIG. 5 is a block diagram of an image processing apparatus according to an exemplary embodiment of the present disclosure;

FIG. 6 is a block diagram of another image processing apparatus shown in an exemplary embodiment of the present disclosure;

FIG. 7 is a schematic structural diagram of an electronic device in an exemplary embodiment of the present disclosure.

Detailed ways

Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. Where the following description refers to the drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the illustrative examples below are not intended to represent all implementations consistent with this disclosure. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present disclosure as recited in the appended claims.

The terminology used in the present disclosure is for the purpose of describing particular embodiments only and is not intended to limit the present disclosure. As used in this disclosure and the appended claims, the singular forms "a," "the," and "the" are intended to include the plural forms as well, unless the context clearly dictates otherwise. It will also be understood that the term "and/or" as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in this disclosure to describe various pieces of information, such information should not be limited by these terms. These terms are only used to distinguish the same type of information from each other. For example, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information, without departing from the scope of the present disclosure. Depending on the context, the word "if" as used herein can be interpreted as "at the time of" or "when" or "in response to determining."

In the process of collecting, transmitting, and processing images, electronic devices are inevitably disturbed by various factors, resulting in the acquired images accompanied by noise signals. The noise signal accompanying the image mainly includes two kinds of noise signal, namely luminance noise and color noise (also known as: chrominance noise).

In the related art, a relatively mature noise reduction technology has been developed for the processing of luminance noise. However, for the color noise in the image, it is still difficult to perform effective noise reduction processing, especially at the color edge with a large color span, color overflow or blurring often occurs.

Specifically, since color noise is more obvious in low-frequency images, related technologies usually use a multi-scale noise reduction framework to perform noise reduction processing on images. In the process of processing through the framework, at least one down-sampling operation needs to be performed on the to-be-processed image to obtain at least one low-frequency image corresponding to the to-be-processed image. Then, noise reduction can be performed on the obtained at least one low-frequency image and the image to be processed, respectively, and then an up-sampling operation is performed on the low-frequency image whose noise reduction has been completed to obtain a final processed image.

Taking two downsampling operations as an example, in this example, the image to be processed is called a high-frequency image, the image obtained by the first downsampling is called an intermediate frequency image, and the image obtained by the second downsampling is called a high-frequency image. for low frequency images. specific:

First, a down-sampling operation can be performed on the high-frequency image to obtain an intermediate-frequency image with a smaller size; and then a down-sampling operation can be performed on the obtained intermediate-frequency image to obtain a low-frequency image with a smaller size. Then, noise reduction processing can be performed on the high frequency image, the intermediate frequency image and the low frequency image respectively, so as to perform relatively detailed noise reduction processing. After the noise reduction process is completed, the low-frequency image can be up-sampled to obtain an image with the same size as the intermediate-frequency image, and the image is used for synthesizing the intermediate-frequency image after the noise reduction process; further, the synthesized image can be An up-sampling operation is performed to obtain an image whose size is consistent with the high-frequency image, and the image is used to synthesize the high-frequency image after noise reduction processing to obtain a processed image.

In the above-mentioned process of noise reduction through the multi-scale noise reduction framework, during the upsampling process of low-frequency images (including the above-mentioned upsampling operation for intermediate frequency images), only upsampling is performed based on its own chromaticity information, and the color cannot be resolved. Color bleeding or blurring at the edges.

More seriously, since the above series of operations of "downsampling operation → noise reduction processing → upsampling operation" involve multiple changes in image scale, the image resolution will be reduced, and this problem may not exist originally. (For example, in the process of upsampling, the determined interpolation is only determined by its own chromaticity value, which may cause the chromaticity difference between adjacent pixels to shrink, resulting in the phenomenon of color blurring at the color edge. ).

To this end, the present disclosure proposes a method that can identify the color edge of an image, and adjust the chromaticity value of the pixel at the color edge to remove the color noise at the color edge, thereby avoiding the color in the related art. Color bleeding or blurring occurs at the edges.

FIG. 1 shows an image processing method according to an exemplary embodiment of the present disclosure. As shown in Figure 1, the method may include the following steps:

Step 102: Acquire an image to be processed in the luminance-chrominance space domain, and perform a downsampling operation on the to-be-processed image to obtain a low-frequency image of the to-be-processed image.

The technical solutions of the present disclosure can be applied to any type of electronic device, for example, the electronic device can be a mobile terminal such as a smart phone and a tablet computer, or a fixed terminal such as a smart TV and a PC (Personal Computer). It should be understood that any electronic device that can perform image processing can be used as the electronic device in the present disclosure. The specific type of electronic device to which the technical solution of the present disclosure is applied can be determined by those skilled in the art according to actual needs. There is no restriction on this disclosure.

It is understandable that the phenomenon of color overflow or blur at the color edge in the image is caused by the insufficient difference of the chromaticity values between the color edge and the color flat place. For example, assuming that the border between the color flat area A and the color flat area B is the color edge C, if the chromaticity value of the pixel X at the color edge is 5, and the color flat area A is closer to the color edge C The chromaticity value of the pixel Y is 4. Obviously, the chromaticity value of the two pixels is not very different, so that C at the color edge is not very obvious, that is, from the visual effect, it is easy to appear color blurred, or the color edge of C is easy to appear. The case where the color spreads to the color flat area A. It can be seen that the reason for the color overflow phenomenon at the color edge in the image is that the difference in pixel chromaticity value between the color edge and the color flat area is small.

In view of this, the present disclosure takes into account the rule of “areas with large color changes in the image, the brightness changes are usually large”, and determines the color edge and the color flat area in the image by the brightness value of each pixel in the image, and then is Pixels in different regions are assigned chroma weights used to adjust chroma values. Among them, the pixels at the color edge are assigned higher chroma weights, and the pixels in the color flat area are assigned relatively lower chroma weights, so that the chroma value of the pixels at the color edge of the adjusted image is the same as the color flat area. The chrominance difference of the pixels in the area is increased, thereby avoiding the problem of color overflow at the color edge.

Since the present disclosure needs to adjust the chrominance value of the pixel based on the luminance value of the pixel, it is obviously required that the image to be processed is an image in the luminance-chrominance space domain. Therefore, when the acquired initial to-be-processed image is an image in another spatial domain, it is necessary to preferentially convert the image in the other spatial domain into an image in the luminance-chrominance spatial domain. For example, when the initial to-be-processed image is an image in the RGB space domain, the image in the RGB space domain should be preferentially converted into an image in the luminance-chrominance space domain.

The to-be-processed image acquired in the present disclosure can be any type of image in the luminance-chrominance space domain. For example, the image to be processed may be an image in the YUV space domain; or, may be an image in the HIS space domain. The specific type of image in the luminance-chrominance space domain of the image to be processed can be determined by those skilled in the art according to the actual situation, which is not limited in the present disclosure.

Since what the present disclosure needs to remove is the color noise in the image to be processed, the color noise usually appears in the form of low frequency. Therefore, after obtaining the to-be-processed image in the luminance-chrominance space domain, the to-be-processed image may be preferentially down-sampled to obtain a low-frequency image of the to-be-processed image, and then the above-mentioned chrominance value adjustment operation is performed on the low-frequency image.

Before introducing the technical solution of the present disclosure in detail, it should be stated that, since this solution involves many concepts, for the convenience of distinction, the pixel in the image to be processed is called "the first pixel", and the pixels in the image to be processed are called "first pixels". Neighborhood pixels are called "first neighborhood pixels", the window area obtained from the image to be processed is called "first window area", and the pixels in the low-frequency image are called "second pixels". The neighborhood pixels of all pixels are called "second neighborhood pixels", and the window area obtained from the low-frequency image is called "second window area".

Actually, in the related art, in addition to the color overflow or blur phenomenon at the color edge of the image, color noise also exists in the color flat area. Therefore, the present disclosure may also include a step of performing noise reduction processing on the image to be processed and/or the low-frequency image. Wherein, when noise reduction processing is performed on the image to be processed, the entire image to be processed can be traversed by means of a sliding window, and the chromaticity values of all the first pixels in any first window area obtained through the sliding window are weighted Calculate, and use the calculated first target value corresponding to any first window area as the chromaticity value of the center pixel of any first window area.

The same is true when noise reduction is performed on low-frequency images. The entire low-frequency image can be traversed by means of a sliding window, and the chromaticity values of all the second pixels in any second window area obtained through the sliding window are weighted and calculated, and the calculated value of any second window is calculated. The second target value corresponding to the region is used as the chromaticity value of the center pixel of any second window region.

In practical operation, most of the image to be processed and the low-frequency image are subjected to noise reduction processing, but it is also possible to perform noise reduction processing on only one of the two. Those skilled in the art can determine whether the image to be processed and/or the low-frequency image is to be processed according to actual needs. The image is subjected to noise reduction processing, which is not limited in the present disclosure.

What needs to be declared is that the sliding window is a more conventional way of taking values in the image field, usually sliding a window with a fixed size at a specific step, so as to use the value of the pixels in the image that falls within the window as the value for operation (including chrominance value, luminance value and other channel values). In the present disclosure, the entire image can be traversed with a step size of 1 pixel, so that each pixel in the image can be used as a central pixel to perform chromaticity value adjustment.

In actual calculation, the chromaticity value of the central pixel of any one of the above-mentioned first window regions can be calculated in various ways.

In one embodiment, after any first window area is obtained from the image to be processed by sliding the window, the difference between all the first pixels in the any first window area and any one of the first window areas can be calculated. The difference value of the center pixel, so that the noise reduction weight is assigned to all the first pixels in the first window area according to the difference value, and the noise reduction weight of any first pixel in the first window area, The difference value corresponding to any one of the first pixels is negatively correlated. On this basis, the chromaticity values of all the first pixels in any first window region can be weighted and averaged based on the noise reduction weight of each first pixel in the first window region, so as to calculate the The obtained value is used as the chromaticity value of the center pixel of any one of the first window regions.

After obtaining any second window area from the low-frequency image by sliding the window, the chromaticity value of the center pixel of the any second window area may be adjusted in a similar manner. Exemplarily, the difference between all the second pixels in the any second window area and the center pixel in the any second window area can be calculated, so that all the first pixels in the any second window area are calculated according to the difference. Two pixels are assigned a noise reduction weight, and the noise reduction weight of any second pixel in the any second window area is negatively correlated with the difference corresponding to the any second pixel. On this basis, the chrominance values of all the second pixels in any second window region can be weighted and averaged based on the noise reduction weight of each second pixel in the second window region, so as to calculate the The obtained value is used as the chromaticity value of the center pixel of any second window area.

For example, assuming that the chromaticity value of each first pixel in the image to be processed is shown in Figure 2, the entire image can be traversed by sliding a window to obtain several first window areas, and the The chromaticity value of each first pixel in the chromaticity value of the central pixel in any first window area is adjusted.

Taking the first window area A shown in FIG. 2 as an example, how to adjust the chromaticity value of the center pixel of any first window area in this embodiment is described:

It can be seen from FIG. 2 that the first window area A includes 9 first pixels, and the chromaticity value of the central pixel X is 102. Then, the noise reduction weights of the nine first pixels in the first window area A are determined based on the chrominance value of 102. Wherein, the noise reduction weight of any first pixel in the first window area A is negatively correlated with the difference between the any first pixel and 102 . For example, the chromaticity value of the first pixel M in the upper left corner of the first window area A is 95, and the difference from 102 is 7; the chromaticity value of the first pixel N in the lower left corner is 85, and the difference from 102 is 85. 17. Since 7 is less than 17, the noise reduction weight of the first pixel M is greater than the noise reduction weight of the first pixel N. For example, the noise reduction weight of the first pixel M may be 1.10, then the noise reduction weight of the first pixel N may be 0.60 . It is assumed that the noise reduction weights of all the first pixels in the first window area A are determined in this way as shown in Table 1 below:

Table 1

Then, on the basis of Table 1, the chrominance value of the center pixel after noise reduction processing can be calculated according to the noise reduction weights of all the first pixels in the first window area A. For example, the calculation can be performed by a weighted average algorithm, and the calculation process can be as follows:

(95*1.1+94*1.0+85*0.60+83*0.55+91*0.80+93*0.95+99*1.4+108*1.2+102*1.45)/(1.1+1.0+0.60+0.55+0.80+0.95 +1.4+1.2+1.45)=96.40.

At this time, 96.40 can be used as the chromaticity value of the central pixel of the first window area A, that is, the original chromaticity value 102 of the central pixel of the first window area A is adjusted to 96.40. It should be understood that, since the sliding window will traverse the entire image, all the first pixels in the image will be used as center pixels, and then go through the above process of adjusting the chrominance value. When the chrominance values of all the first pixels are adjusted, it can be considered that the noise reduction processing of the graphics to be processed is completed.

For the process of performing chromaticity adjustment on any second pixel in the low-frequency image, refer to the above example, and the processing methods for the “low-frequency image”, “second pixel” and “second window area” can be adapted to refer to the “image to be processed” The processing methods of the "first pixel" and the "first window area" will not be repeated here.

It can be seen from the above adjustment process that the larger the chroma value difference between the sliding window and the center pixel, the smaller the noise reduction weight, so that the adjusted chroma value of the center pixel is different from the chroma value of the neighboring pixels. value shrinks. From the perspective of visual effect, the color noise in the image will be filtered out, and the noise reduction effect for the color noise will be realized.

However, as can be seen from Table 1 above, since the above embodiment determines the noise reduction weight of each first pixel in the first window area A based on the chromaticity value of the central pixel, the noise reduction weight of the central pixel itself is always the largest, which leads to Effective noise reduction is difficult to achieve when denoising isolated colored noise in color flat areas. Take the first window area B in FIG. 2 as an example: it is not difficult to see that in the first window area B, the chromaticity value of the central pixel is 125, and the chromaticity value is much larger than that of other first pixels in the first window area B. The chromaticity value, and the chromaticity values between several first neighborhood pixels around the central pixel are relatively close. Obviously, visually, the center pixel belongs to the isolated color noise in the color flat area. In the above embodiment, the noise reduction weight of the central pixel is always the largest, so that after the noise reduction processing is performed on the first pixel with a chromaticity value of 125, the chromaticity value of the central pixel is still much larger than that of the first pixel of the central pixel. The chrominance value of a neighborhood pixel, that is, the isolated noise in the color flat area cannot be effectively removed.

In view of this, the present disclosure proposes another noise reduction processing method.

In another embodiment, the noise reduction weight of all pixels in any window area is no longer determined based on the central pixel, but is based on the mean value of the chromaticity values of all pixels in any window area, as the Pixels in either window region are assigned denoising weights.

For example, in the process of performing noise reduction processing on the image to be processed, the following operations may be performed on any first window area: The first difference between the chromaticity value of each first pixel in the any first window area and the chromaticity mean value determines the first noise reduction weight of each first pixel in the any first window area, wherein , the first noise reduction weight of any first pixel in the any first window area is negatively correlated with the first difference value corresponding to the any first pixel; then, based on the any first window The first noise reduction weight of each first pixel in the area, and the weighted average calculation of the chromaticity values of all the first pixels in the first window area is performed to obtain the first target value corresponding to the any first window area, and use the first target value as the chromaticity value of the center pixel of any one of the first window regions.

In the process of noise reduction processing for low-frequency images, the following operations can be performed on any second window area: first, calculate the chromaticity mean of all the second pixels in the According to the second difference between the chromaticity value of each second pixel in the any second window area and the chromaticity mean value, the second noise reduction weight of each second pixel in the any second window area is determined, Wherein, the second noise reduction weight of any second pixel in the any second window area is negatively correlated with the second difference value corresponding to the any second pixel; The second noise reduction weight of each second pixel in the window area, the weighted average calculation of the chromaticity values of all the second pixels in the second window area is performed, and the second target value corresponding to the second window area is obtained. , and the second target value is taken as the chromaticity value of the center pixel of any second window area.

Taking the first window area B in the image to be processed shown in FIG. 2 as an example, how to adjust the chromaticity value of the center pixel of any first window area in this embodiment is described:

First calculate the mean chromaticity of all the first pixels in the first window area B: (98+86+99+89+85+91+66+84+125)/9=91.44 (for the convenience of calculation, 91 is used in the following text example). Then, calculate the first difference between the chromaticity value of each first pixel in the first window area B and the average chromaticity value, so as to assign the drop value to the corresponding first pixel according to the first difference value corresponding to the first pixel of each pixel. noise weight. For example, the assigned noise reduction weights can be as shown in Table 2 below:

Table 2

Then, on the basis of Table 2, the chrominance value of the center pixel after noise reduction processing can be calculated according to the noise reduction weights of all the first pixels in the first window area B. For example, the calculation can be performed by a weighted average algorithm, and the calculation process can be as follows:

(98*1.1+86*1.3+99*1.0+89*1.35+85*1.20+91*1.45+66*0.40+84*1.10+125*0.30)/(1.1+1.3+1.0+1.35+1.20+1.45 +0.40+1.10+0.30)=90.11.

At this time, 90.11 can be used as the chromaticity value of the central pixel, that is, the original chromaticity value 125 of the central pixel of the first window area B is adjusted to 90.11. As in the previous embodiment, since the sliding window will traverse the entire image to be processed, all the first pixels in the image will be used as center pixels, and then go through the above process of adjusting the chrominance value. When the chrominance values of all the first pixels are adjusted, it can be considered that the noise reduction processing of the image to be processed is completed.

For the process of performing chromaticity adjustment on any second pixel in the low-frequency image in this embodiment, reference may also be made to the above example, and the processing methods for the "low-frequency image", "second pixel" and "second window area" can be adapted to the reference " The processing methods of "image to be processed", "first pixel" and "first window area" will not be repeated here.

It can be seen from the above Table 2 that in this embodiment, the chromaticity value of the central pixel of 125 is no longer used as the basis for assigning weights to each first pixel in the first window area B, so that the first pixel with the largest weight is usually not For example, in the above table 2, since the obtained chromaticity mean value is 91, it is obvious that in the window area B, the first pixel with a pixel value of 91 has the largest noise reduction weight, thereby avoiding the previous embodiment. , because the weight value of the central pixel is always the largest, resulting in a technical problem that the isolated color noise in the color flat area cannot be removed.

It is not difficult to understand that, in the previous embodiment, since the noise reduction weight is assigned to each pixel based on the center pixel, the noise reduction weight of the center pixel is always the largest. Obviously, if the central pixel is an isolated color noise in a flat color area (that is, the chromaticity value is quite different from the chromaticity value of the surrounding pixels), after adjusting the chromaticity value of the central pixel on this basis, the The chrominance value is still quite different from the chrominance value of the surrounding pixels, so that the color noise cannot be removed.

However, in this embodiment, since the weight is assigned based on the chromaticity mean value of all pixels in the window area, the noise reduction weight of the central pixel is usually not the maximum. On the contrary, if the central pixel is an isolated color noise in a flat color area, since the chromaticity values of multiple neighboring pixels of the central pixel are not much different, the obtained chromaticity mean value will be the same as the central pixel's chromaticity value. The chrominance difference value is large, but the chrominance difference value from the neighboring pixels of the central pixel is small, so that the noise reduction weight of the central pixel is relatively small. For example, in the first window area B of the above example, the central Pixel denoising weights the least. Obviously, after noise reduction processing is performed on the image in this embodiment, the isolated color noise in the color flat area can be effectively eliminated. For example, after adjusting the chromaticity value of the central pixel of the window area B, the adjusted chromaticity value of 90.11 is obviously close to the chromaticity value of most of the first pixels in the window area B. From the visual effect, it is There is no isolated color noise in the color flat areas.

It should be stated that the numerical values of the noise reduction weights of the first pixels in the above Tables 1 and 2 are only schematic, and are only used to represent: the noise reduction weight of any first pixel is the same as that of any first pixel. The difference with the center pixel (or the chrominance mean above) is negatively correlated. How to specifically determine the noise reduction weight of each first pixel can be determined by those skilled in the art according to the actual situation, which is not limited in the present disclosure. For example, the corresponding relationship between the difference value and the noise reduction weight can be set as the basis for determining the noise reduction weight of any first pixel; for example, each first pixel can be sorted according to the difference value corresponding to each first pixel , to assign noise reduction weights according to the order of arrangement. The same is true when noise reduction processing is performed on low-frequency images, and details are not repeated here.

Step 104: Determine the chrominance weight of each first pixel according to the brightness difference between each first pixel in the to-be-processed image and the respective first neighborhood pixels, wherein the chrominance weight of each first pixel is , the chrominance weight of each first pixel is positively correlated with the corresponding luminance difference degree.

It can be seen from the above content that the phenomenon of color overflow or blur at the color edge in the image is caused by the too small difference between the chromaticity value of the pixel at the color edge and the chromaticity value of the pixel in the color flat area. Therefore, in the present disclosure, color edges and color flat areas in the image are determined based on the luminance values of each pixel in the image, and a higher chrominance weight is assigned to the pixels at the color edge, and pixels in the color flat area are assigned higher chrominance weights. Assign relatively lower chroma weights to increase the difference between the chroma values of the pixels at the edges of the color and the chroma values of the pixels in the flat areas of the color, thereby solving the problem of color bleeding or blurring at the edges of the color.

It should be stated that, after the low-frequency image is obtained by downsampling the image to be processed, the present disclosure assigns chrominance weights to each first pixel in the image to be processed according to the luminance value of each first pixel in the image to be processed. Then, the chromaticity value of each second pixel in the low-frequency image is adjusted in the process of upsampling the low-frequency image with the help of the correspondence between each first pixel in the to-be-processed image and each second pixel in the low-frequency image. In other words, the chromaticity value of each second pixel in the low-frequency image is adjusted based on the luminance distribution in the image to be processed.

The reason for this is that the low-frequency image is obtained based on the image to be processed, and the brightness distribution of the two is almost the same, and the image to be processed can be regarded as: compared with the high-frequency image of the low-frequency image, the size is larger and contains more brightness details. , the chrominance weight of each first pixel is determined by the image to be processed, which is more accurate.

Step 106: Perform an up-sampling operation on the low-frequency image based on the chrominance weight corresponding to each pixel, and synthesize the image obtained through the up-sampling operation with the to-be-processed image to obtain a processed image.

In actual operation, the first neighborhood pixels of each first pixel can be determined by sliding the window, so as to determine the chromaticity of each first pixel according to the brightness difference between each first pixel and the respective first neighborhood pixels Weights. Taking the first window area A shown in FIG. 2 as an example, other first pixels except the center pixel can be regarded as the first neighborhood pixels of the center pixel (of course, in the calculation process, the center pixel can also be regarded as a its own first neighborhood pixels). Of course, the first window area A only takes the sliding window including 9 first pixels as an example. If the sliding window includes 25 first pixels, the number of first neighborhood pixels also increases relatively. The specific manner of determining the pixels in the first neighborhood can be determined by those skilled in the art according to actual needs, which is not limited in the present disclosure.

In one embodiment, when assigning a chrominance weight to any one of the first pixels, the assignment may be performed in combination with all the first neighborhood pixels of the any one of the first pixels. For example, the luminance difference value between any first pixel and any first neighborhood pixel may be preferentially calculated, and a weight parameter corresponding to any first neighborhood pixel may be calculated based on the luminance difference value. After the weight parameters corresponding to all the first neighborhood pixels of the any first pixel are calculated by this method, the chrominance weight can be assigned to the any first pixel based on all the calculated weight parameters.

Still taking the first window area A shown in FIG. 2 as an example, assuming that the value of each first pixel shown in FIG. 2 is the luminance value of the corresponding first pixel, then, for any The following calculation is performed for one pixel: take the first pixel L in the upper right corner with a luminance value of 99 as an example (it is a first neighborhood pixel of the center pixel X in the first window area A, which can be regarded as any of the above-mentioned first neighborhoods pixel), calculate the luminance difference value between the first pixel L and the central pixel X to be 3, and assign a weight parameter to the first pixel L based on the luminance difference value of 3. And after completing the above operations on all the first neighborhood pixels of the center pixel X in the first window area A, after obtaining the weight parameters of each first neighborhood pixel corresponding to the center pixel X, all the obtained weight parameters can be obtained. , determine the chrominance weight of the central pixel X in the first window area A.

It should be noted that, in this embodiment, although the luminance difference between any first pixel and its first neighboring pixels is used, each first neighboring pixel is assigned a weight parameter. However, in actual operation, any pixel can be regarded as its own neighborhood pixel, that is, a weight parameter is also assigned to any first pixel itself. In other words, when the weight parameters are allocated to the first neighboring pixels of the central pixel X, 9 weight parameters are finally allocated.

After obtaining the chrominance weights of all the first pixels in the to-be-processed image in the above manner, the chrominance value of each second pixel in the low-frequency image can be adjusted during the upsampling operation on the low-frequency image.

For example, when calculating the chromaticity value of any pixel in the image obtained by the upsampling operation, the luminance value of the first pixel corresponding to the any pixel in the image to be processed, and the low-frequency image corresponding to the any pixel can be calculated. The chromaticity value of the second pixel corresponding to one pixel is used as a basis for calculation. Since the size of the image obtained by the upsampling operation is the same as the size of the image to be processed, it can also be expressed as: according to the brightness value of any first pixel in the image to be processed, the color of the second pixel corresponding to any pixel in the low-frequency image The chromaticity value of the target pixel corresponding to any one of the first pixels in the image obtained through the upsampling operation is calculated.

In actual operation, after obtaining the weight parameter of any first neighborhood pixel of any first pixel in the to-be-processed image by the above method, the second pixel corresponding to the any first pixel can be determined in the low-frequency image. pixel, and further determine a second neighborhood pixel corresponding to any one of the first neighborhood pixels from the second neighborhood pixels of the second pixel. On this basis, the distance between the second neighborhood pixel and the second pixel can be calculated, and based on the distance and the weight parameter of any first neighborhood pixel, determine the distance between the second neighborhood pixel and any first neighborhood pixel. Corresponding chromaticity reference value; after determining a plurality of chromaticity reference values corresponding to all the first neighborhood pixels of the first pixel, the weighted average calculation can be performed on the plurality of chromaticity reference values to obtain Obtain the chrominance value of the target pixel corresponding to any one of the first pixels in the image obtained through the upsampling operation.

For example, to calculate the chrominance value of any pixel in the image obtained by the upsampling operation, the following formula can be referred to:

为通过上采样操作得到的图像中的任一像素的色度值；p为待处理图像中与上述任一像素对应的第一像素，q为该第一像素的任一第一邻域像素；I_p为第一像素p的亮度值，I_q为第一邻域像素q的亮度值；p_↓为第一像素p在低频图像中对应的第二像素(实际为该第二像素的坐标，为方便后续描述，用第二像素p_↓表示该第二像素)，q_↓为第一邻域像素q在低频图像中对应的第二邻域像素(实际为该第二像素的坐标，为方便后续描述，用第二像素q_↓表示该第二像素)；‖p_↓-q_↓‖为第二邻域像素q_↓与第二像素p_↓的距离；

为第一像素p在低频图像中对应的第二像素的色度值；Ω为以第二像素p_↓为中心像素的第二窗口区域；k_p为归一化常量；f()为根据距离计算色度参考值的系数函数；g(‖I_p-I_q‖)为根据亮度差值计算权重参数的权重函数，用于计算上述权重参数，

为上述色度参考值。in,

is the chromaticity value of any pixel in the image obtained by the upsampling operation; p is the first pixel corresponding to any of the above-mentioned pixels in the image to be processed, and q is any first neighborhood pixel of the first pixel; I _p is the brightness value of the first pixel p, I _q is the brightness value of the first neighborhood pixel q; p _↓ is the second pixel corresponding to the first pixel p in the low-frequency image (actually the coordinates of the second pixel, For the convenience of subsequent description, the second pixel p _↓ is used to represent the second pixel), and q _↓ is the second neighborhood pixel corresponding to the first neighborhood pixel q in the low-frequency image (actually the coordinates of the second pixel, for convenience In the following description, the second pixel q _↓ is used to represent the second pixel); ‖p _↓ -q _↓ ‖ is the distance between the second neighborhood pixel q _↓ and the second pixel p _↓ ;

is the chromaticity value of the second pixel corresponding to the first pixel p in the low-frequency image; Ω is the second window area with the second pixel p _↓ as the center pixel; k _p is a normalized constant; The coefficient function for calculating the chrominance reference value; g(‖I _p -I _q ‖) is the weight function for calculating the weight parameter according to the luminance difference value, which is used to calculate the above weight parameter,

is the above-mentioned chromaticity reference value.

For ease of understanding, the image to be processed shown in FIG. 2 is still taken as an example, and it is assumed that each numerical value in FIG. 2 is a brightness value corresponding to each first pixel. It is assumed that a low-frequency image obtained by down-sampling operation through the chromaticity diagram corresponding to FIG. 2 is shown in FIG. 3 . Obviously, there is a certain correspondence between the first pixel in FIG. 2 and the second pixel in FIG. 3 . For example, the size of the image to be processed shown in FIG. 2 is 8*8, while the size of the low-frequency image shown in FIG. 3 is 4*4, which means that any second pixel in the low-frequency image corresponds to a pixel in the image to be processed. 4 first pixels, and the corresponding relationship is related to the position of the pixel in the image. For example, the second pixel with the upper left corner chromaticity value of 56 in the low-frequency image should correspond to the four first pixels with the upper left corner luminance values of 65, 56, 77, and 58 in the image to be processed.

What is currently known is that the size of the image obtained by performing the upsampling operation on the low-frequency image shown in FIG. 3 must be as shown in FIG. 4A . So in the upsampling process, all that needs to be done is how to know the chrominance value of each pixel in the image shown in Figure 4A. Taking the chromaticity value of the pixel X" in Fig. 4A as an example for introduction, it is obvious that the pixel X" corresponds to the central pixel of the first window area A in Fig. 2 (ie, the first pixel X in Fig. 2), and The first pixel X belongs to one of the four pixels in the upper right corner in FIG. 2 , then the second pixel corresponding to the first pixel X in FIG. 3 is the second pixel X′ with a chromaticity value of 91 in the upper right corner in FIG. 3 . . Similar to the first window area A, in the low-frequency image, the second window area with the second pixel X' as the center pixel can be determined by sliding the window. Assuming that a weight parameter needs to be assigned to the first neighborhood pixel Y of the first pixel X, the weight parameter of the first neighborhood pixel Y can be determined based on the luminance difference between the first pixel X and the first neighborhood pixel Y. Determining the weight parameters of other neighborhood pixels of the pixel X is also similar; after determining the weight parameters of the first neighborhood pixel Y, it can be determined from several second neighborhood pixels of the second pixel X' that are related to the first neighborhood pixel. The second neighborhood pixel Y' corresponding to Y, and then the chromaticity of the first neighborhood pixel Y is determined based on the distance between the second neighborhood pixel Y' and the second pixel X' and the weight parameter of the first neighborhood pixel Y The reference value (of course, it can also be regarded as the chrominance reference value of the second neighborhood pixel Y'). The formula specifically used to calculate the chromaticity reference value can be determined by those skilled in the art according to actual needs, which is not limited in the present disclosure.

It should be stated that, in the second window area corresponding to the second pixel X' shown in FIG. 3 , the area beyond the low-frequency image is included. In this partial area, the chromaticity value of the second pixel (which may be called a blank pixel) that does not exist in the original image can be supplemented by means of complement value. In the art, the complementary value technology is relatively mature, and any method can be used to add complementary value to the blank pixels in the second window area, which is not limited in the present disclosure. For example, for any blank pixel, the chromaticity value of the second pixel of the closest blank pixel may be used as its chromaticity value.

After the chromaticity reference values of all the first neighborhood pixels of the first pixel X are obtained in this way, the weighted average calculation of the plurality of chromaticity reference values can be performed to obtain the color of the pixel X” corresponding to the first pixel X. degree value.

To reflect the difference between the present disclosure and the related art, reference may be made to FIG. 4B . Figure (a) in Figure 4B can be regarded as a chromaticity diagram of the image to be processed, Figure (b) is a luminance diagram of the image to be processed, Figure (c) is a low-frequency image obtained by downsampling Figure (a), (d) is a processed image obtained by up-sampling by bilinear interpolation in the related art, and Figure (e) is a processed image obtained by up-sampling by the method of the present disclosure. Since the present disclosure utilizes the luminance map of the image to be processed to upsample the low-frequency image, this upsampling operation may also be referred to as joint guided upsampling.

Comparing Figure (d) and Figure (e): In Figure (e) obtained by joint guided upsampling, the pixel with a chroma value of 136 (the dark 136 in Figure (d)) has a pixel in its left neighborhood. The chromaticity value is 132, and the chromaticity difference between the two is 4; in the figure (d) obtained by upsampling through bilinear interpolation, the chromaticity value of the pixel corresponding to the above-mentioned "pixel with a chromaticity value of 136" is 135, the chromaticity value of the domain pixel on the left is 133, and the difference between the two is 2. Obviously, the processed image obtained by joint guided upsampling has a larger chromaticity difference between the pixels at the color edge, which can improve the chromaticity difference between the pixels at the color edge compared with the related art. value to avoid color bleeding and blurring at the edges of the color.

It can be seen from the above description that, in this embodiment, the determination of the chrominance weight and the adjustment of the chrominance value through the chrominance weight are completed in the process of performing the upsampling operation. From the computing point of view of the electronic device, the above-mentioned multiple steps are completed by one formula. Of course, in addition to completing the above steps in the process of performing the upsampling operation on the low-frequency image, the above steps can also be completed in a step-by-step manner.

In another embodiment, in the to-be-processed image, the average luminance of any first pixel and all its first neighboring pixels, and the second difference between the luminance value of any first pixel and the average luminance may be calculated preferentially ; and then determine the chrominance weight corresponding to any one of the first pixels based on the second difference value, for example, the product of the second difference value and a predefined coefficient can be used as the chrominance weight of any one of the first pixels. Based on the same method, the chrominance weights of all the first pixels in the image to be processed can be obtained.

Correspondingly, a conventional up-sampling operation can also be performed on the low-frequency image. For example, the low-frequency image may be calculated by an up-sampling interpolation algorithm to obtain an image to be weighted that is consistent with the size of the image to be processed. The up-sampling interpolation algorithm may be any type of interpolation algorithm, for example, a two-dimensional interpolation algorithm and a linear interpolation algorithm may be used as the interpolation algorithm used in this embodiment. After the image to be weighted is obtained, the corresponding pixels in the image to be weighted can be weighted based on the chrominance weight of each first pixel in the image to be processed, and the image obtained by the weighted calculation is used as the image obtained through the upsampling operation. image.

Still taking the to-be-processed image shown in FIG. 2 as an example, assuming that the chromaticity weight of the first pixel X needs to be determined at present, the average brightness of all the first pixels in the first window area A can be calculated, and the brightness value can be obtained: (95+94+85+83+91+93+99+108+102)/9=94.4, then the calculated second difference corresponding to the first pixel X is 7.6. The second difference value 7.6 determines the chrominance weight of the first pixel X. In the same way, the chrominance weights of other first pixels in the image to be processed can be obtained.

On the other hand, it is still assumed that the image shown in FIG. 3 is a low-frequency image obtained by down-sampling the chromaticity diagram of the image to be processed shown in FIG. 2 , wherein the numerical values shown in FIG. The chrominance value of each second pixel. In this embodiment, a conventional up-sampling operation can be performed on the low-frequency image to obtain an image to be weighted with the same size as the image to be processed. For example, the image to be weighted can be as shown in FIG. 4A . The chromaticity value of each pixel is not shown in FIG. 4A , but the chromaticity value of each pixel of the image to be weighted obtained in this embodiment is determined. On this basis, the chrominance value of the corresponding pixels in the to-be-weighted image can be adjusted based on the chrominance weight of each first pixel in the image to be processed. For example, based on the chrominance weight of the first pixel X, the ” to adjust the chroma value.

It can be seen from the above technical solutions that, on the one hand, the present disclosure performs a downsampling operation on the obtained image to be processed in the luminance-chrominance space domain to obtain a corresponding low-frequency image; on the other hand, according to each first pixel in the image to be processed A chrominance weight is assigned to each of the included first pixels with respect to the luminance difference degree of the respective first neighborhood pixels, wherein the chrominance weight of each first pixel is positively correlated with the corresponding luminance difference degree. On this basis, an up-sampling operation can be performed on the low-frequency image based on the chrominance weight corresponding to each first pixel in the image to be processed to obtain an image with the same size as the image to be processed, and the image and the image to be processed are merged , to get the processed image.

It should be understood that, in the field of images, the change of color in the image will have an impact on the change of brightness, and the area where the color change is larger usually has a larger change in the brightness. Therefore, for any pixel, if the degree of brightness difference with neighboring pixels is greater, it means that any pixel is in an area with a large color change, that is, at the color edge. In the present disclosure, the chrominance weight of any pixel is positively correlated with the corresponding chrominance weight, which is equivalent to assigning a higher chrominance weight to a pixel in an area with a large luminance change (at the color edge). On this basis, up-sampling the low-frequency image by the determined chrominance weight is equivalent to increasing the chroma value of the pixel at the color edge in the low-frequency image, making the color change at the color edge more prominent, thereby avoiding The problem of color overflow at the color edge in the related art.

In short, the present disclosure identifies the color edge in the image by analyzing the brightness change in the image, and then assigns a higher chrominance weight to the pixel at the color edge, so that the pixel at the color edge The difference between the chromaticity value of the pixel and the chromaticity value of the pixel where the color is flat increases, which highlights the difference between the color edge and the color flat, and solves the problem of color overflow at the color edge in the related art.

Optionally, the present disclosure may also perform noise reduction processing on the image to be processed and/or the low-frequency image. For example, the image to be processed and/or the low-frequency image can be traversed by means of a sliding window, and the chromaticity values of all pixels in any window area obtained through the sliding window are weighted and calculated, so as to use the obtained value as the The chrominance value of the center pixel of any window area. Wherein, in the process of weighting calculation, a noise reduction weight can be assigned to each pixel according to the mean of all pixels in any window area, and the noise reduction weight of any pixel is negatively correlated with the difference between the pixel and the above mean.

Since the mean value of all pixels in any window area is used as a standard to assign a noise reduction weight to each pixel, the noise reduction weight of the central pixel is not necessarily the largest. Among them, when the isolated color noise in the color flat area is denoised, the chromaticity difference between the central pixel corresponding to the color noise and the neighboring pixels is quite different, resulting in the obtained chromaticity mean and the neighboring pixels. The difference between the chrominance value is small, and the difference from the chrominance value of the central pixel is relatively large, which will make the noise reduction weight of the central pixel minimum. It can be seen that the above method can effectively reduce the difference between the chromaticity value of the pixel corresponding to the color noise and the chromaticity value of the adjacent pixels, thereby effectively removing the isolated color noise in the color flat area.

Corresponding to the foregoing embodiments of the image processing method, the present disclosure also provides embodiments of an image processing apparatus.

FIG. 5 is a block diagram of an image processing apparatus according to an exemplary embodiment of the present disclosure. Referring to FIG. 5 , the apparatus includes an acquisition unit 501 , a determination unit 502 and a synthesis unit 503 .

The acquiring unit 501 acquires an image to be processed in the luminance-chrominance space domain, and performs a downsampling operation on the image to be processed to obtain a low-frequency image of the image to be processed;

The determining unit 502 determines the chrominance weight of each first pixel according to the brightness difference between each first pixel in the to-be-processed image and the respective first neighborhood pixels, wherein each first pixel is , the chrominance weight of each first pixel is positively correlated with the corresponding degree of luminance difference;

The synthesis unit 503 performs an up-sampling operation on the low-frequency image based on the chrominance weight corresponding to each first pixel, and synthesizes the image obtained through the up-sampling operation with the to-be-processed image to obtain a processed image.

Optionally, the determining unit 502 is further assembled as:

Calculate the luminance difference between any first pixel and any first neighborhood pixel, and calculate a weight parameter corresponding to the any first neighborhood pixel based on the luminance difference;

The chrominance weight of any one of the first pixels is determined based on the calculated weight parameters of all the first neighborhood pixels of the any one of the first pixels.

Optionally, the synthesis unit 503 is further assembled as:

Perform the following operations on all the first pixels in the to-be-processed image: acquire a second pixel corresponding to any first pixel in the low-frequency image, and determine among all the second neighborhood pixels of the second pixel, a second neighborhood pixel corresponding to any of the first neighborhood pixels;

Calculate the distance between the second neighborhood pixel and the second pixel, and determine the pixel corresponding to any of the first neighborhood pixels based on the distance and the weight parameter of any of the first neighborhood pixels The chromaticity reference value of ;

Perform a weighted average calculation on the determined multiple chromaticity reference values of all the first neighborhood pixels corresponding to the any first pixel, to obtain the chromaticity value of the target pixel corresponding to the any first pixel;

After obtaining the chromaticity values of all target pixels corresponding to all the first pixels in the image to be processed, an image obtained through the upsampling operation is generated based on the chromaticity values of all the target pixels.

Optionally, the determining unit 502 is further assembled as:

calculating the luminance mean value of any first pixel and all its first neighboring pixels, and the second difference between the luminance value of any first pixel and the luminance mean value;

The product of the second difference and a predefined coefficient is used as the chrominance weight of any one of the first pixels.

Optionally, the synthesis unit 503 is further assembled as:

Calculate the low-frequency image through an upsampling interpolation algorithm to obtain an image to be weighted that is consistent with the size of the image to be processed;

Based on the chromaticity weight of each first pixel in the to-be-processed image, weighted calculation is performed on the chromaticity value of the corresponding pixel in the to-be-weighted image;

The image obtained by the weighting calculation is regarded as the image obtained through the upsampling operation.

As shown in FIG. 6 , FIG. 6 is a block diagram of another image processing apparatus shown in an exemplary embodiment of the present disclosure. On the basis of the foregoing embodiment shown in FIG. 5 , this embodiment further includes: a noise reduction unit 504 .

Optionally, also include:

The noise reduction unit 504 is configured to traverse the to-be-processed image through a sliding window, and perform weighted calculation on the chromaticity values of all the first pixels in any first window area obtained through the sliding window, to obtain the the first target value corresponding to any first window area; taking the first target value as the chromaticity value of the center pixel of the any first window area; and/or, traversing the low frequency by means of a sliding window image, and perform weighted calculation on the chromaticity values of all the second pixels in any second window area obtained through the sliding window to obtain the second target value corresponding to the any second window area; The target value is used as the chromaticity value of the center pixel of any second window area.

Optionally, the noise reduction unit 504 is further assembled as:

Calculate the chromaticity mean value of all the first pixels in the any first window area, and according to the first difference between the chromaticity value of each first pixel in the any one window area and the chromaticity mean value, Determine the first noise reduction weight of each first pixel; the noise reduction weight of any first pixel in any of the first window regions is negative with the first difference corresponding to the any first pixel correlation; based on the noise reduction weight of each first pixel in the any first window area, perform a weighted average calculation on the chromaticity values of all the first pixels in the any first window area, and obtain the any first window area. a first target value corresponding to a window area; and/or,

Calculate the chromaticity mean value of all the second pixels in the any second window area, and according to the second difference between the chromaticity value of each second pixel in the any second window area and the chromaticity mean value value, determine the second noise reduction weight of each second pixel; the noise reduction weight of any second pixel in any second window area, and the second difference value corresponding to the any second pixel is negatively correlated; based on the noise reduction weight of each second pixel in the any second window area, the weighted average calculation is performed on the chromaticity values of all the second pixels in the any second window area, and the arbitrary second window area is obtained. A second target value corresponding to a second window area.

Regarding the apparatus in the above-mentioned embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be described in detail here.

For the apparatus embodiments, since they basically correspond to the method embodiments, reference may be made to the partial descriptions of the method embodiments for related parts. The device embodiments described above are only illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed over multiple network elements. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the present disclosure. Those of ordinary skill in the art can understand and implement it without creative effort.

Correspondingly, the present disclosure also provides an image processing apparatus, comprising: a processor; a memory for storing instructions executable by the processor; wherein the processor is configured to implement the image according to any one of the foregoing embodiments A processing method, for example, the method may include: acquiring an image to be processed in the luminance-chrominance space domain, and performing a downsampling operation on the image to be processed to obtain a low-frequency image of the image to be processed; The degree of brightness difference between each first pixel and the respective first neighborhood pixels, respectively determine the chrominance weight of each first pixel, wherein, in each first pixel, the chrominance weight of each first pixel It is positively correlated with the corresponding brightness difference; based on the chrominance weight corresponding to each first pixel, an upsampling operation is performed on the low-frequency image, and the image obtained through the upsampling operation is synthesized with the to-be-processed image, Get the processed image.

Correspondingly, the present disclosure also provides an electronic device comprising a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by one or more processors Executing the one or more programs includes instructions for implementing the image processing method described in any of the foregoing embodiments. For example, the method may include: acquiring a to-be-processed image in the luminance-chrominance space domain, and applying the Perform a downsampling operation on the image to be processed to obtain a low-frequency image of the image to be processed; according to the degree of difference in brightness between each first pixel in the to-be-processed image and the respective first neighborhood pixels, determine the respective first pixels The chrominance weight of the pixel, wherein, among the first pixels, the chrominance weight of each first pixel is positively correlated with the corresponding brightness difference degree; An up-sampling operation is performed on the image, and the image obtained through the up-sampling operation is synthesized with the to-be-processed image to obtain a processed image. FIG. 7 is a block diagram of an apparatus 700 for implementing a process scheduling method according to an exemplary embodiment. For example, apparatus 700 may be a mobile phone, computer, digital broadcast terminal, messaging device, game console, tablet device, medical device, fitness device, personal digital assistant, and the like.

7, the apparatus 700 may include one or more of the following components: a processing component 702, a memory 704, a power supply component 706, a multimedia component 708, an audio component 710, an input/output (I/O) interface 712, a sensor component 714, and communication component 716 .

The processing component 702 generally controls the overall operation of the device 700, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 702 can include one or more processors 720 to execute instructions to perform all or some of the steps of the methods described above. Additionally, processing component 702 may include one or more modules to facilitate interaction between processing component 702 and other components. For example, processing component 702 may include a multimedia module to facilitate interaction between multimedia component 708 and processing component 702.

Memory 704 is configured to store various types of data to support operations at device 700 . Examples of such data include instructions for any application or method operating on device 700, contact data, phonebook data, messages, pictures, videos, and the like. Memory 704 may be implemented by any type of volatile or nonvolatile storage device or combination thereof, such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic or Optical Disk.

Power supply assembly 706 provides power to the various components of device 700 . Power components 706 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to device 700 .

Multimedia component 708 includes screens that provide an output interface between the device 700 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touch, swipe, and gestures on the touch panel. The touch sensor may not only sense the boundaries of a touch or swipe action, but also detect the duration and pressure associated with the touch or swipe action. In some embodiments, multimedia component 708 includes a front-facing camera and/or a rear-facing camera. When the apparatus 700 is in an operation mode, such as a shooting mode or a video mode, the front camera and/or the rear camera may receive external multimedia data. Each of the front and rear cameras can be a fixed optical lens system or have focal length and optical zoom capability.

Audio component 710 is configured to output and/or input audio signals. For example, audio component 710 includes a microphone (MIC) that is configured to receive external audio signals when device 700 is in operating modes, such as calling mode, recording mode, and voice recognition mode. The received audio signal may be further stored in memory 704 or transmitted via communication component 716 . In some embodiments, audio component 710 also includes a speaker for outputting audio signals.

The I/O interface 712 provides an interface between the processing component 702 and a peripheral interface module, which may be a keyboard, a click wheel, a button, or the like. These buttons may include, but are not limited to: home button, volume buttons, start button, and lock button.

Sensor assembly 714 includes one or more sensors for providing status assessment of various aspects of device 700 . For example, the sensor assembly 714 can detect the open/closed state of the device 700, the relative positioning of components, such as the display and keypad of the device 700, and the sensor assembly 714 can also detect a change in the position of the device 700 or a component of the device 700 , the presence or absence of user contact with the device 700 , the orientation or acceleration/deceleration of the device 700 and the temperature change of the device 700 . Sensor assembly 714 may include a proximity sensor configured to detect the presence of nearby objects in the absence of any physical contact. Sensor assembly 714 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 714 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

Communication component 716 is configured to facilitate wired or wireless communication between apparatus 700 and other devices. The apparatus 700 may access a wireless network based on a communication standard, such as WiFi, 2G or 3G, 4G LTE, 5G NR (New Radio), or a combination thereof. In one exemplary embodiment, the communication component 716 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 716 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

In an exemplary embodiment, apparatus 700 may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable A gate array (FPGA), controller, microcontroller, microprocessor or other electronic component implementation is used to perform the above method.

In an exemplary embodiment, there is also provided a non-transitory computer-readable storage medium including instructions, such as a memory 704 including instructions, executable by the processor 720 of the apparatus 700 to perform the method described above. For example, the non-transitory computer-readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like.

Other embodiments of the present disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common general knowledge or techniques in the technical field not disclosed by this disclosure . The specification and examples are to be regarded as exemplary only, with the true scope and spirit of the disclosure being indicated by the following claims.

It is to be understood that the present disclosure is not limited to the precise structures described above and illustrated in the accompanying drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

The above descriptions are only preferred embodiments of the present disclosure, and are not intended to limit the present disclosure. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present disclosure shall be included in the present disclosure. within the scope of protection.

Claims (10)

1. An image processing method, characterized by comprising:

acquiring a to-be-processed image of a brightness-chrominance space domain, and performing down-sampling operation on the to-be-processed image to obtain a low-frequency image of the to-be-processed image;

respectively determining the chromaticity weight of each first pixel according to the brightness difference degree of each first pixel and each first neighborhood pixel in the image to be processed, wherein the chromaticity weight of each first pixel in each first pixel is positively correlated with the corresponding brightness difference degree;

and performing upsampling operation on the low-frequency image based on the chromaticity weight corresponding to each first pixel, and synthesizing the image obtained through the upsampling operation with the image to be processed to obtain a processed image.

2. The method of claim 1, further comprising, prior to the upsampling operation on the low frequency image:

traversing the image to be processed in a sliding window mode, and performing weighted calculation on chrominance values of all first pixels in any first window area acquired through the sliding window to obtain a first target value corresponding to any first window area; taking the first target value as a chrominance value of a center pixel of any one of the first window regions; and/or the presence of a gas in the gas,

traversing the low-frequency image in a sliding window mode, and performing weighted calculation on chrominance values of all second pixels in any second window area acquired through the sliding window to obtain a second target value corresponding to any second window area; the second target value is taken as the chrominance value of the center pixel of said any second window area.

3. The method of claim 2,

the obtaining a first target value corresponding to any window region by performing weighted calculation on chrominance values of all first pixels in any first window region acquired through a sliding window includes:

calculating the chrominance mean value of all the first pixels in any one of the first window regions, and determining a first noise reduction weight of each first pixel according to a first difference value between the chrominance value of each first pixel in any one of the first window regions and the chrominance mean value; the noise reduction weight of any first pixel in any first window region is in negative correlation with a first difference value corresponding to any first pixel; based on the noise reduction weight of each first pixel in any first window region, performing weighted average calculation on chrominance values of all first pixels in any first window region to obtain a first target value corresponding to any first window region;

the performing weighted calculation on the chrominance values of all second pixels in any second window region acquired through the sliding window to obtain a second target value corresponding to any second window region includes:

calculating the chrominance mean value of all second pixels in any second window region, and determining a second noise reduction weight of each second pixel according to a second difference value between the chrominance mean value and each second pixel in any second window region; the noise reduction weight of any second pixel in any second window region is in negative correlation with a second difference value corresponding to any second pixel; and performing weighted average calculation on the chrominance values of all the second pixels in any second window region based on the noise reduction weight of each second pixel in any second window region to obtain a second target value corresponding to any second window region.

4. The method of claim 1, wherein determining the chroma weight of each first pixel according to the brightness difference degree between the first pixel and the first neighboring pixel in the image to be processed comprises:

calculating the brightness difference value of any first pixel and any first neighborhood pixel thereof, and calculating a weight parameter corresponding to any first neighborhood pixel based on the brightness difference value;

and determining the chrominance weight of any first pixel based on the calculated weight parameters of all first neighborhood pixels of the any first pixel.

5. The method of claim 4, wherein the upsampling the low-frequency image based on the chrominance weights corresponding to the first pixels comprises:

performing the following operations on all first pixels in the image to be processed: acquiring a second pixel corresponding to any first pixel in the low-frequency image, and determining a second neighborhood pixel corresponding to any first neighborhood pixel in all second neighborhood pixels of the second pixel;

calculating the distance between the second neighborhood pixels and the second pixels, and determining a chromaticity reference value corresponding to any one first neighborhood pixel based on the distance and the weight parameter of any one first neighborhood pixel;

performing weighted average calculation on a plurality of determined chromaticity reference values of all first neighborhood pixels corresponding to any one first pixel to obtain a chromaticity value of a target pixel corresponding to any one first pixel;

after obtaining the chrominance values of all target pixels corresponding to all first pixels in the image to be processed, generating an image obtained through the upsampling operation based on the chrominance values of all the target pixels.

6. The method of claim 1, wherein determining the chrominance weight of each first pixel according to the degree of the luminance difference between the first pixel and the respective first neighboring pixel in the image to be processed comprises:

calculating the brightness mean value of any first pixel and all first neighborhood pixels thereof and a second difference value of the brightness value of any first pixel and the brightness mean value;

and taking the product of the second difference value and a predefined coefficient as the chroma weight of any first pixel.

7. The method of claim 1, wherein the upsampling the low-frequency image based on the chrominance weights corresponding to the first pixels comprises:

calculating the low-frequency image through an up-sampling interpolation algorithm to obtain an image to be weighted with the size consistent with that of the image to be processed;

based on the chroma weight of each first pixel in the image to be processed, carrying out weighted calculation on the chroma value of the corresponding pixel in the image to be weighted;

and taking the image obtained by the weighting calculation as the image obtained by the up-sampling operation.

8. An image processing apparatus characterized by comprising:

the device comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring an image to be processed in a brightness-chromaticity space domain and performing down-sampling operation on the image to be processed to obtain a low-frequency image of the image to be processed;

the determining unit is used for respectively determining the chromaticity weight of each first pixel according to the brightness difference degree of each first pixel and each first neighborhood pixel in the image to be processed, wherein the chromaticity weight of each first pixel in each first pixel is positively correlated with the corresponding brightness difference degree;

and the synthesizing unit is used for performing up-sampling operation on the low-frequency image based on the chrominance weight corresponding to each first pixel, and synthesizing the image obtained through the up-sampling operation with the image to be processed to obtain a processed image.

9. An electronic device, comprising:

a processor;

a memory for storing processor-executable instructions;

wherein the processor implements the method of any one of claims 1-7 by executing the executable instructions.

10. A computer-readable storage medium having stored thereon computer instructions, which when executed by a processor, perform the steps of the method according to any one of claims 1-7.

Images