CN100420279C - Method and device for reducing image noise under low illumination - Google Patents

Abstract

The invention provides an image processing method and device for reducing image noise under low illumination so as to improve image quality. The method includes the following steps: determining whether the shooting environment is a dark place; if so, switching the output mode of the image from a color image to a grayscale image; and performing low-pass filtering on the switched grayscale image.

Description

Method and device for reducing image noise under low illumination

field of invention

The invention relates to a digital image processing system, in particular to a method and system for reducing noise and improving image quality under relatively low illuminance.

Background technique

Existing color camera equipment generally has an automatic exposure adjustment function, so that the equipment can automatically maintain a stable image brightness output as the ambient light intensity changes. As the light intensity decreases, the image processing system adopted by the device prolongs the exposure time of the photosensitive device, and in some cases increases the brightness gain accordingly. The result of this is that although the output amplitude of the photosensitive device is small, the image brightness can still be basically maintained at the same level after a long exposure time and the adjustment of the brightness gain.

When the ambient illuminance drops below a certain critical value, such as at night, the image noise increases obviously. There are two main reasons for this. One is that when the photosensitive device is in the photosensitive state for a long time, due to the influence of white noise, dark current, etc., the output signal already contains a lot of noise components, and these noises are very random. Second, when the ambient brightness is relatively dark, the system has to adjust the brightness gain to a large value, which in turn amplifies the noise in the output signal of the photosensitive device. Due to the influence of these reasons, the image contains a lot of noise, which is more obvious in the color image. As a result, color image quality is poor in low-light environments.

Contents of the invention

The object of the present invention is to provide an image processing method and device that overcome the above-mentioned problems.

According to a first aspect of the present invention, there is provided a method for reducing image noise in an image processing system, the method comprising the steps of: determining whether the shooting environment is a dark place; if so, switching the output mode of the image from a color image to a grayscale image ; and, perform low-pass filtering on the switched grayscale image.

In the first aspect, preferably, the step of determining whether the shooting environment is a dark place includes the step of detecting whether the exposure time of the image processing system is greater than a first predetermined value.

Preferably, the step of determining whether the shooting environment is a dark place includes the step of detecting whether the brightness gain coefficient of the image processing system is greater than a second predetermined value.

Preferably, the switching step includes averaging the R, G, and B values of each pixel as the R, G, and B values of the corresponding pixel in the output image.

Preferably, the switching step includes calculating the Y value according to the following formula for the R, G, and B values of each pixel: Y=0.299R+0.587G+0.114B, and then using the Y value as the corresponding value in the output image The R, G, and B values of the pixel.

Preferably, the step of switching includes calculating the X value according to the following formula for the R, G, and B values of each pixel: X=max(R, G, B), and then using the X value as the corresponding value in the output image The R, G, and B values of the pixel.

Preferably, the low-pass filtering is performed by a Gaussian low-pass filter.

According to a second aspect of the present invention, there is provided an image processing system, the system comprising: a device for determining whether the shooting environment is a dark place; when the determining device determines that the shooting environment is a dark place, the output mode of the image is switched from a color image a grayscale switching device for a grayscale image; and a filtering device for performing low-pass filtering on the switched grayscale image.

In the second aspect, preferably, the means for determining whether the shooting environment is a dark place detects whether the exposure time of the image processing system is greater than a first predetermined value.

Preferably, the device for determining whether the shooting environment is a dark place detects whether the brightness gain coefficient of the image processing system is greater than a second predetermined value.

Preferably, the grayscale switching device averages the R, G, and B values of each pixel as the R, G, and B values of the corresponding pixels in the output image.

Preferably, the grayscale switching device calculates the Y value according to the following formula for the R, G, and B values of each pixel: Y=0.299R+0.587G+0.114B, and then uses the Y value as the corresponding value in the output image The R, G, and B values of the pixel.

Preferably, the grayscale switching device calculates the X value according to the following formula for the R, G, and B values of each pixel: X=max(R, G, B), and then uses the X value as the corresponding value in the output image. The R, G, and B values of the pixel.

Preferably, said filtering means is a Gaussian low-pass filter.

The present invention switches the color image into a grayscale image, so the mutual influence brought by the noise of each color component is eliminated. At the same time, low-pass filtering the gray image also reduces the noise of the image, further improving the quality of the image.

Description of drawings

The present invention will be described in more detail below with reference to accompanying drawing, wherein:

Fig. 1 is the method for reducing image noise under low illumination environment according to the present invention; With

FIG. 2 is a schematic block diagram of the structure of the image processing system of the present invention.

Detailed ways

FIG. 1 schematically shows a method for reducing image noise in a low-light environment according to the present invention. As shown in FIG. 1 , at step 110 , the exposure time or brightness gain coefficient of the system is firstly calculated. Then, enter step 120, judge whether exposure time is greater than a certain predetermined value T, or whether the brightness gain coefficient is greater than a certain value Gmax; . Of course, the user can also directly set the current shooting environment as a dark place.

Then, the program goes to step 140 . In this step, the output mode of the image is switched from color image to grayscale image. In the grayscale image, the interaction caused by the noise of each color component is avoided. For example, many red, blue and other noise points randomly appear on the image of the face, so that the image does not look very messy. In addition, in a very dark environment, the dependence of human visual perception on the color of the object is greatly reduced, and the perception of the object is mainly based on contrast, outline, relative motion, etc., so it is completely unnecessary to preserve the color image.

There are three specific implementation methods for switching from a color image to a grayscale image:

Method 1: Average the R, G, and B values of each pixel as the R, G, and B values of the corresponding pixels in the output image. In this way, the values of the three color channels are the same, which is equivalent to a grayscale image.

Method 2: For the R, G, and B values of each pixel, calculate the Y value according to the following formula:

Y=0.299R+0.587G+0.114B.

Then, take the Y value as the R, G, and B values of the corresponding pixel in the output image.

Method 3: For the R, G, and B values of each pixel, calculate the X value according to the following formula:

X = max(R, G, B).

Then, use the X value as the R, G, B value of the corresponding pixel in the output image.

Certainly, other forms of image switching methods are also feasible.

Next, the program low-pass filters the image at step 150 . Low-pass filtering can suppress the influence of random noise to a certain extent. Because the distribution of random noise in the spectrum is relatively uniform, and the spectrum distribution of normal signals has a certain width, so through the low-frequency filter, the spectral components of normal signals can be better preserved and a large part of random noise can be removed. Since the grayscale image is filtered, the filter can be designed relatively simply. For example, the filter may be a Gaussian low-pass filter.

FIG. 2 is a schematic block diagram of the structure of the image processing system of the present invention. As shown in FIG. 2 , the digital photosensitive device 210 generates images (images) under the action of light, and the generated images are input to the digital gain adjustment device 220 and the brightness statistics device 230 .

The brightness statistics device 230 makes statistics on the brightness of the image to obtain a statistical value of the image, such as the average brightness. The average luminance is sent to exposure time calculation means 240 and gain coefficient calculation means 250 .

The exposure time calculation device 240 calculates an appropriate exposure time value according to the input brightness value, and sends the value to the digital photosensitive device 210 to control the exposure time of the device 210. This value is also sent to the low illumination determination device 260 at the same time.

The gain coefficient calculation device 250 calculates an appropriate gain coefficient value according to the input brightness value, and sends the value to the digital gain device 220 and the low-illuminance determination device 260 .

The digital gain adjustment device 220 performs gain adjustment on the image according to the input gain coefficient, and the image after gain adjustment is input to the grayscale switching device 270 .

The low illumination determining means 260 determines whether the current shooting environment is a dark place according to the input exposure time value or gain coefficient value. If the exposure time is greater than a certain predetermined value T, or whether the brightness gain coefficient is greater than a certain value Gmax, it is determined that the current shooting environment is a dark place. Of course, the user can also directly set the current shooting environment as a dark place.

The grayscale switching device 260 switches the input image into a grayscale image. The specific switching method is as described above. Then the switched grayscale image is sent to the low-pass filtering device 280 for low-pass filtering.

The present invention is mainly applied to occasions including but not limited to mobile phone cameras, PC cameras and digital cameras.

While the invention has been described in some detail, such description is illustrative only and not restrictive. It will be obvious to a person skilled in the art that the invention can be varied in many ways.

For example, each of the devices 220-280 may be implemented in a monolithic integrated circuit, or may be implemented as discrete circuits or devices.

All such variations are within the scope of the invention and should not be considered as a departure from the scope of the invention. The invention is to be limited only by the spirit and scope of the appended claims.

Claims (14)

1. A method for reducing image noise in an image processing system, the method comprising the steps of: Determine whether the shooting environment is a dark place; If so, switch the output mode of the image from a color image to a grayscale image; and Low-pass filter the switched grayscale image. 2. The method according to claim 1, wherein the step of determining whether the shooting environment is a dark place comprises the step of detecting whether the exposure time of the image processing system is greater than a first predetermined value. 3. The method according to claim 1, wherein the step of determining whether the shooting environment is a dark place comprises the step of detecting whether the brightness gain coefficient of the image processing system is greater than a second predetermined value. 4. The method according to any one of claims 1-3, wherein the step of switching the output mode of the image from a color image to a grayscale image comprises averaging the R, G, and B values of each pixel as the R, G, and B values of the corresponding pixel in the output image. 5. The method according to any one of claims 1-3, characterized in that the step of switching the output mode of the image from a color image to a grayscale image comprises R, G, and B values of each pixel according to the following formula Find the Y value: Y=0.299R+0.587G+0.114B, and then use the Y value as the R, G, and B values of the corresponding pixel in the output image. 6. The method according to any one of claims 1-3, wherein the step of switching the output mode of the image from a color image to a grayscale image comprises the following formula for the R, G, and B values of each pixel Find the X value: X=max(R, G, B), and then use the X value as the R, G, and B values of the corresponding pixel in the output image. 7. The method according to any one of claims 1-3, characterized in that the low-pass filter is completed by a Gaussian low-pass filter. 8. An image processing system, the system comprising: A device to determine whether the shooting environment is a dark place; When the determination device determines that the shooting environment is a dark place, a grayscale switching device for switching the output mode of the image from a color image to a grayscale image; and A filtering device for performing low-pass filtering on the switched grayscale image. 9. The image processing system according to claim 8, wherein the device for determining whether the shooting environment is a dark place detects whether the exposure time of the image processing system is greater than a first predetermined value. 10. The image processing system according to claim 8, wherein the device for determining whether the shooting environment is a dark place detects whether the brightness gain coefficient of the image processing system is greater than a second predetermined value. 11. The image processing system according to any one of claims 8-10, wherein the grayscale switching device averages the R, G, and B values of each pixel as the R, G values of the corresponding pixels in the output image. , B value. 12. The image processing system according to any one of claims 8-10, wherein the grayscale switching device calculates the Y value for each pixel's R, G, and B values according to the following formula: Y=0.299R+ 0.587G+0.114B, and then use the Y value as the R, G, and B values of the corresponding pixel in the output image. 13. The image processing system according to any one of claims 8-10, characterized in that the grayscale switching device calculates the X value according to the following formula for the R, G, and B values of each pixel: X=max(R , G, B), and then use the X value as the R, G, B value of the corresponding pixel in the output image. 14. The image processing system according to any one of claims 8-10, wherein the filtering device is a Gaussian low-pass filter.

Images