KR102771405B1 - Optical element arrangement method of surface lighting device and sample image correction method of surface defect inspection device - Google Patents

Abstract

The present invention provides a method for arranging optical elements in a surface illuminator of a surface defect inspection device, wherein a plurality of optical elements are arranged in a square array inside a housing of a surface illuminator, and the spacing between the optical elements is adjusted based on the shadows appearing in an image of a subject being inspected, so that images of surface defects of various types, shapes, and directions can be significantly distinguished as the shadows are corrected, thereby greatly improving the reliability of detection for various surface defects of the subject being inspected, and further, a lookup table is configured to calculate a correction value inversely, and the sample image is corrected using the Gaussian and Lambertian effects, thereby eliminating brightness unevenness of a sample image due to the Gaussian and Lambertian effects, and a sample image correction method of the surface defect inspection device.

Description

Optical element arrangement method of surface lighting device and sample image correction method of surface defect inspection device

The present invention relates to a method for arranging optical elements in a surface illuminator of a surface defect inspection device and a method for correcting a sample image of the surface defect inspection device. More specifically, the present invention relates to a method for arranging optical elements in a surface illuminator of a surface defect inspection device, which arranges a plurality of optical elements in a rectangular array inside a housing of the surface illuminator and adjusts the spacing between the optical elements based on the shadow appearing in an image of a subject to be inspected, thereby correcting the shadow, and a method for correcting a sample image of the surface defect inspection device, which configures a lookup table with Gaussian and Lambertian effects and inversely calculates a correction value to correct the image.

In general, for products in the form of flat plates, such as steel plates, PCB substrates, semiconductor wafers, and films, a surface inspection process is conducted to evaluate whether there are any defects such as contaminants, scratches, dents, or cracks on the surface of the product after the manufacturing process is completed.

In the past, the surface inspection process was usually carried out by having an inspector directly inspect the surface of a product using the naked eye or a microscope, but in the case of such visual inspection, since the presence or absence of a defect is determined by the inspector's subjective judgment based on visual observation, it was difficult to secure reliability for uniform quality, and there was a problem that it could not be applied to surface inspection of secondary batteries, semiconductor wafers, PCB substrates, etc., where minute surface defects can have a fatal impact on quality.

In order to solve these problems, a surface inspection device has recently been developed that determines whether or not a defect exists on the surface of a product by using a light source that irradiates light onto the surface of the product to be inspected and an image sensor that captures an image of the reflected light reflected from the surface of the product. This surface inspection device using an image sensor is disclosed in detail in prior art, Korean Patent No. 10-0389967 (February 27, 2002).

However, in the case of surface defects, there is a problem in that it is difficult to accurately detect surface defects of various types, shapes, or directions when applying a surface illumination method that uniformly incidents light on the inspection surface, as in the prior art, because the type, shape, and direction of the defect that can be detected through the output of the image sensor differs depending on the direction in which light is incident.

For this reason, in recent years, surface lighting has been incident on the surface of the test object at various angles. However, when surface lighting is incident at various angles, uneven areas are created on the incident surface due to the occurrence of shadows, and a lighting structure to solve this problem has been required.

The present invention provides a method for arranging optical elements in a surface illuminator of a surface defect inspection device, which arranges a plurality of optical elements in a square array inside a housing of the surface illuminator, and adjusts the spacing between the optical elements based on the shadows appearing in an image of a subject being inspected, so that images of surface defects of various types, shapes, and directions can be significantly distinguished as the shadows are corrected, thereby greatly improving the reliability of detection of various surface defects of a subject being inspected, and a method for correcting a sample image of the surface defect inspection device.

The method for arranging optical elements provided in a surface illuminator of a surface defect inspection device according to the present invention relates to a method for arranging optical elements provided in a surface illuminator of a surface defect inspection device, which comprises a frame portion providing a transport path for an object to be inspected, a transport portion installed on one side of the frame portion and transporting the object to be inspected along the transport path, an optical module support portion installed on one side of the frame portion in the middle of the transport path, and a surface illuminator and an image sensor portion fixedly installed on the optical module support, the method comprising the steps of: a) arbitrarily arranging a plurality of optical elements in a rectangular array inside a rectangular housing of the surface illuminator; b) applying power to the optical elements to surface illuminate the object to be inspected with the surface illuminator, while acquiring a rectangular image corresponding to an area of the housing; c) measuring an X-axis direction shading of the image based on the acquired rectangular image; d) a step of grouping optical elements in the Y-direction array among the plurality of optical elements arranged in a square shape into one Y-direction array based on the measured X-direction shading, and then moving the grouped Y-direction arrays to adjust the X-direction spacing of the optical elements; e) a step of measuring Y-direction shading of an image based on the acquired rectangular image; and f) a step of grouping optical elements in the X-direction array among the plurality of optical elements arranged in a square shape into one X-direction array based on the measured Y-direction shading, and then moving the grouped X-direction arrays to adjust the Y-direction spacing of the optical elements.

At this time, the square arrangement of the optical elements according to the present invention is such that a plurality of optical elements are arranged at regular intervals along the X-axis direction of the rectangular housing to form a plurality of X-axis-direction columns (horizontal columns), a plurality of optical elements are arranged at regular intervals along the Y-axis direction to form a plurality of Y-axis-direction columns, and by combining the plurality of X-axis-direction columns and the plurality of Y-axis-direction columns, the plurality of optical elements are respectively arranged at the vertices of the square.

And in the step c) of measuring the X-axis direction shading of the image based on the acquired rectangular image according to the present invention, the X-axis direction length of the shading is measured based on the X-axis direction of the image.

In addition, in the step d) according to the present invention, based on the measured X-axis shading, among the plurality of square-arranged optical elements, the Y-axis-arranged optical elements are each grouped into one Y-axis-direction array, and then the grouped Y-axis-direction arrays are moved to adjust the X-axis-direction spacing of the optical elements, it is preferable to arrange the Y-axis-direction arrays so that the spacing between them becomes narrower as they go toward the shading side.

In addition, in the step of measuring the Y-axis direction shading of the image based on the acquired rectangular image in step e) according to the present invention, the Y-axis direction length of the shading is measured based on the Y-axis direction of the image.

At this time, based on the measured Y-axis shading in the step f) according to the present invention, among the plurality of square-arranged optical elements, the optical elements in the X-axis array are each grouped into one X-axis array, and then the grouped X-axis arrays are moved to adjust the Y-axis spacing of the optical elements. In this step, it is preferable that the spacing of the X-axis arrays grouped in the X-axis direction be arranged so that the spacing becomes narrower as they go toward the shading side.

In addition, in the step b) according to the present invention, in which power is applied to the optical elements to illuminate the surface of the subject with a surface illuminator and a rectangular image corresponding to the housing area is obtained, it is preferable to place an illuminance meter at each corner position of the subject area to measure the illuminance of the incident area, adjust the light source to achieve uniform illuminance over the entire subject area, and then obtain an image of the subject illuminated with the surface illumination.

The image correction method of a surface defect inspection device according to the present invention comprises the steps of: i) taking multiple pictures of the subject to be inspected while sequentially incident light from different directions based on an incident area of light among the subject to be inspected, thereby obtaining sample images corresponding to different incident angles of light; ii) optimizing the brightness of each of the obtained sample images; iii) optimizing the incident angles of the brightness-optimized sample images; and iv) obtaining a corrected final sample image.

At this time, in the step i) according to the present invention, in which the test subject is photographed multiple times while sequentially incident light from different directions based on the light incidence area among the test subject, and each sample image corresponding to a different light incidence angle is acquired, information on the incidence angle and brightness of the light is collected together when each sample image is acquired.

And in the step ii) according to the present invention, the step of optimizing the brightness of the acquired sample image comprises the steps of ii-1) extracting an average pixel value among the pixels forming the sample image, ii-2) searching for a relatively brightest pixel among the pixels forming the sample image and extracting a max pixel value based on the gray level of the searched pixel, ii-3) searching for a relatively darkest pixel among the pixels forming the sample image and extracting a min pixel value based on the gray level of the searched pixel, ii-4) calculating a Wmax pixel value, which is a weight for adjusting the extracted max pixel value to the average pixel value, and calculating a Wmin pixel value, which is a weight for adjusting the extracted min pixel value to the average pixel value, ii-5) calculating a distance (Distance) between the darkest pixel (min pixel value) based on the brightest pixel (max pixel value), and ii-6) calculating a change in weight per unit pixel distance (Kstep). The method includes: a step of calculating a weight (K) of each pixel forming a sample image; and a step of ii-8) optimizing the brightness of the sample image by applying the calculated weight to each pixel forming the sample image.

Here, the change in weight per unit pixel distance (Kstep) according to the present invention is calculated by the following [Mathematical Formula 1].

[Mathematical Formula 1]

Kstep=(Wmin-Wmax)/Dmax

(Here, Kstep is the change in weight per unit pixel distance, Wmin is the weight for the min pixel value to become the average pixel value, Wmax is the weight for the max pixel value to become the average pixel value, and Dmax is the distance between the darkest pixel and the brightest pixel.)

In addition, the weight (K) of each pixel according to the present invention is calculated by the following [Mathematical Formula 2].

[Mathematical formula 2]

K= Wmax+Kstep×Distance

(Here, K is the weight of each pixel, Wmax is the weight for the max pixel value to become the average pixel value, Kstep is the change in weight per unit pixel distance, and Distance is the distance between each pixel based on the brightest pixel.)

And in the step iii) according to the present invention, the step of performing the incident angle optimization of the sample images for which the brightness has been optimized includes the steps of iii-1) preparing at least three sample images having different incident angles of light from among the sample images for which the brightness has been optimized, iii-2) measuring the actual incident angles of light of each sample image and confirming the incident angles for each pixel of the sample images, iii-3) using the photomatrice stereo modeling to derive a pixel normal vector using three or more types of gray levels and the actual angles of the corresponding pixel based on the actual incident angles obtained for each pixel, and iii-4) using the photomatrice stereo model to input the pixel normal vector and arbitrarily set incident angles of light, thereby obtaining a sample image for which the incident angle has been optimized with a new light brightness.

At this time, based on the actual incidence angle obtained for each pixel in step ⅲ-3) according to the present invention, in the step of calculating the pixel normal vector using three or more types of gray levels and actual angles of the corresponding pixel using photomatrices stereo modeling, the pixel normal vector is calculated by [Mathematical Formula 3] below.

[Mathematical Formula 3]

(Here, ρ is the albedo constant, N is the normal vector, S is the inverse matrix of the light 3-way matrix, and I is the brightness values of each pixel of the three sample images.)

In addition, in the step of obtaining a sample image in which the incident angle is optimized with a new light brightness by inputting the pixel normal vector and the arbitrarily set incident angles of light using the photomatrice stereo model of the step iii-4) according to the present invention,

The incident angle is corrected by [Mathematical Formula 4] below.

[Mathematical formula 4]

,

,

,

,

,

,

(Here, ρ is the albedo constant, N is the normal vector, S is the inverse matrix of each light 3-way matrix of the three sample images, and I is the brightness values of each pixel of the three sample images.)

The effects exhibited by the optical element arrangement method provided in the surface illuminator of the surface defect inspection device according to the present invention and the sample image correction method of the surface defect inspection device are as follows.

By arranging a plurality of optical elements in a square array inside the housing of a surface illuminator, and adjusting the spacing between the optical elements based on the shadows appearing in the image of the subject being inspected, images of surface defects of various types, shapes, and directions can be significantly distinguished as the shadows are corrected, thereby greatly improving the detection reliability of various surface defects of the subject being inspected.

In addition, by configuring the Gaussian and Lambertian effects as a lookup table, calculating the correction value inversely, and using this to correct the sample image, the brightness unevenness of the sample image due to the Gaussian and Lambertian effects can be eliminated.

FIG. 1 is an exemplary diagram showing a surface defect inspection device according to one embodiment of the present invention.
Figure 2 is an exemplary diagram showing the main parts of a surface defect inspection device according to one embodiment of the present invention.
FIG. 3 is an exemplary diagram showing the first to fourth surface illuminators of a surface defect inspection device according to one embodiment of the present invention.
FIG. 4 is an exemplary diagram showing a surface illuminator of a surface defect inspection device according to one embodiment of the present invention.
FIG. 5 is an exemplary diagram showing a state in which shade occurs in an image captured by an image generating unit according to one embodiment of the present invention.
FIG. 6 is an exemplary diagram showing a state in which square-arranged optical elements are bundled into a Y-axis array according to one embodiment of the present invention.
FIG. 7 is an exemplary diagram showing a state in which the spacing between Y-axis direction arrays is adjusted according to one embodiment of the present invention.
FIG. 8 is an exemplary diagram showing a state in which square-arranged optical elements are bundled into an X-axis-direction array according to one embodiment of the present invention.
FIG. 9 is an exemplary diagram showing a state in which the spacing between X-axis direction arrays is adjusted according to one embodiment of the present invention.
FIG. 10 is a block diagram showing a sample image correction method of a surface defect inspection device according to an embodiment of the present invention step by step.
FIG. 11 is a block diagram showing a step-by-step process of optimizing the brightness of a sample image according to one embodiment of the present invention.
FIG. 12 is a block diagram showing a step-by-step process for optimizing the angle of incidence of a sample image according to one embodiment of the present invention.
FIG. 13 is a photograph showing optimized sample images before brightness correction, after brightness correction, and after brightness correction according to one embodiment of the present invention.
FIG. 14 is a photograph showing an optimized sample image before, after, and after correction of the angle of incidence according to one embodiment of the present invention.
Figure 15 is an exemplary diagram showing a measurement state for an incident angle according to one embodiment of the present invention.
FIG. 16 is a photograph showing a sample image corrected before brightness and angle of incidence correction, after brightness and angle difference correction, and after brightness and angle of incidence correction according to one embodiment of the present invention.

Hereinafter, a preferred embodiment according to the present invention will be described in detail with reference to the attached drawings. Prior to this, the terms or words used in this specification and claims should not be interpreted as limited to their usual or dictionary meanings, and should be interpreted as meanings and concepts that conform to the technical idea of the present invention based on the principle that the inventor can appropriately define the concept of the term in order to explain his or her own invention in the best way.

Therefore, the embodiments described in this specification and the configurations illustrated in the drawings are only the most preferred embodiments of the present invention and do not represent all of the technical ideas of the present invention. Therefore, it should be understood that there may be equivalent modified examples that can replace them at the time of filing this application.

The present invention relates to a method for arranging optical elements in a surface illuminator of a surface defect inspection device, which arranges a plurality of optical elements in a square array inside a housing of the surface illuminator, and adjusts the spacing between the optical elements based on the shadow appearing in an image of a subject to be inspected, thereby correcting the shadow. The method will be described with reference to the drawings as follows.

First, a surface defect inspection device according to an embodiment of the present invention with reference to FIGS. 1 and 4 will be examined as follows.

The above surface defect inspection device (1) includes a frame part (10) that provides a transport path for an inspection object, a transport part (30) that is installed on one side of the frame part (10) and transports the inspection object along the transport path, an optical module support (20) that is installed on one side of the frame part (10) in the middle of the transport path, a plurality of surface illuminators (40, 50, 60, 70) that are fixedly installed on the optical module support (20), and an image generation part (80).

Here, the frame portion (10) may be configured in a table shape with the upper surface parallel to the ground, and an opening may be formed in the center of the upper surface to provide a transport path along which the test object is transported along the longitudinal direction of the upper surface.

In addition, the transfer unit (30) performs the function of transferring the test subject along the longitudinal direction of the upper surface of the frame unit (10) (i.e., the transfer path), and the transfer unit (30) includes a screw-shaped transfer shaft (31) installed in the central portion of the upper surface of the frame unit (10), a transfer plate (32) in which a connecting portion on one side of the lower surface is connected to the transfer shaft (31) through a connecting hole in the center and on which the test subject is placed on the upper surface, and a transfer motor (33) that rotates the transfer shaft (31).

At this time, it is preferable that bearings (not shown) are installed on both sides of the length direction of the frame portion (10) to which both ends of the transfer shaft (31) are connected so that the transfer shaft (31) can be rotatably fixed.

Accordingly, the transfer part (30) rotates the transfer shaft (31) coupled in the longitudinal direction to the central portion of the upper surface opening of the frame part (10) by the transfer motor (33), and the transfer plate (32) coupled to the transfer shaft (31) moves along the longitudinal direction of the upper surface of the frame part (10), thereby causing the test object placed on the upper surface of the transfer plate (32) to be transferred along the transfer path.

In the present invention, the above-mentioned transport unit (30) is described as a rotating screw type by a transport motor (33), but is not limited thereto, and any of the known transport means, such as a conveyor belt, may be used within the scope of performing the same function.

In addition, although the frame part (10) and the transport part (30) are shown to transport a single test object, the frame part (10) and the transport part (30) may be configured to transport a plurality of test objects continuously while being spaced apart from each other in order to speed up the surface defect inspection process and improve work efficiency.

And the optical module support (20) is formed in a roughly 'ㄷ' shape in which the upper ends of vertical supports (not shown) on both sides are connected by horizontal supports (not shown), and each end of the vertical supports (not shown) is configured to be fixedly installed at both ends in the width direction of the upper surface of the frame portion (10).

Meanwhile, the image generating unit (80) is fixed to the center of the horizontal support (not shown) of the optical module support (20) so as to be positioned in the middle of the transport path, and captures the surface of the subject being transported along the transport path from above, thereby obtaining an image accordingly.

At this time, it is preferable that the image generation unit (80) uses a photographing device such as a CCD or digital camera, and photographs the entire width direction of the transported test object at each shooting.

In addition, the image generating unit (80) may be configured to continuously take pictures of test objects that are continuously supplied while being spaced apart from each other according to a predetermined shooting cycle, but if necessary, a proximity sensor (not shown) or the like may be used to configure the image generating unit (80) to take pictures according to the shooting cycle from the time when one end of the test object, which is the subject of the examination, approaches the area where the picture is taken by the image generating unit (80) until the other end passes the area where the picture is taken.

At this time, the shooting area, which is the area of the surface of the test subject that is captured by the image generating unit (80), may vary depending on the characteristics of the shooting device. However, it is explained that the shooting area is a rectangular area that has a length in the direction of the transport path and a length corresponding to the width of the test subject (or the width of the transport plate) in the width direction of the frame unit (10).

The above surface illuminators (40, 50, 60, 70) are provided in multiple numbers (total of 4) and each emit light at different locations with respect to the imaging area.

A surface defect inspection device according to one embodiment of the present invention is equipped with a total of four surface illuminators (40, 50, 60, 70), and it is preferable that each of the first to fourth surface illuminators (40, 50, 60, 70) be positioned at a square corner position relative to the inspection object.

More specifically, it includes a first side illuminator (40) including a plurality of optical elements that incident light onto the imaging area from the left and right sides of the front side of the image generating unit (80), a third side illuminator (60), a second side illuminator (50) including a plurality of optical elements that incident light onto the imaging area from the left and right sides of the rear side of the image generating unit (80), a fourth side illuminator (70), and a lighting controller that lights up a plurality of optical elements (42, 52, 62, 72) included in the first side illuminator (40) to the fourth side illuminator (700) in a predetermined order.

At this time, the first side illuminator (40) and the third side illuminator (60) have a length corresponding to the width, and are provided so that the light-emitting surfaces on the left and right sides in front of the image generating unit (80) can be rotated 0° to 90° to selectively incline the direction of the imaging area, and the second side illuminator (50) and the fourth side illuminator (70) also have a length corresponding to the width, and are provided so that the light-emitting surfaces on the left and right sides in the rear of the image generating unit (80) can be rotated 0° to 90° to selectively incline the direction of the imaging area.

Accordingly, the first side illuminator (40) to the fourth side illuminator (70) are respectively positioned at corner points of the imaging area forming a square and spaced apart from each other by a certain distance, and it is preferable that the first side illuminator (40) to the fourth side illuminator (70) be symmetrical with respect to the center of the spacing distance, and each end of both sides is connected to each other by a pair of connecting plates (34), and each connecting plate (34) is fixed to a vertical support (not shown) of the optical module support (20) by a pair of support plates, thereby being installed in the middle of the transport path.

In addition, the lighting controller can be installed on one side of the optical module support (20), and is configured to be fixedly installed on one side of the vertical support (not shown).

When photographing the surface of a subject to be examined, each of the first to fourth surface illuminators (40) to (70) can be rotated at a corresponding angle (typically 45°) relative to the horizontal line or the horizontal line, thereby changing the incident angle of each of the first to fourth surface illuminators (40) to (70) to an oblique line, thereby illuminating the surface of the subject to be examined.

In this case, the vertical incidence angle, which is the angle between the direction in which light is incident on the photographing area and the direction in which the image generating unit (80) photographs the surface of the subject, of the first side illuminator (40) and the third side illuminator (60), becomes different from each other, and also, for the same reason, the vertical incidence angles of the second side illuminator (50) and the fourth side illuminator (70) become different from each other.

In this way, if the directions in which the first to fourth surface illuminators (40) to (70) illuminate the photographing area are different from each other, shadows may occur in the image obtained from the image generating unit (80), which may significantly reduce the reliability of detection of surface defects.

Therefore, in the present invention, the spacing between the optical elements (42, 52, 62, 72) of each of the first side illuminator (40) to fourth side illuminator (70) is adjusted for the purpose of correcting the image shading so that shading does not appear in the image generated by the image generating unit (80).

Referring to FIGS. 3 to 9, the optical element arrangement method provided in the surface illuminator (40, 50, 60, 70) of the surface defect inspection device according to the above-described configuration is as follows.

First, in step a), a plurality of optical elements (42, 52, 62, 72) are arbitrarily arranged in a rectangular array inside a rectangular housing (41, 51, 61, 71) of a surface illuminator (40, 50, 60, 70).

Preferably, a plurality of optical elements (42, 52, 62, 72) are arranged so that uniform surface incidence occurs when the light is vertically incident.

At this time, the square arrangement of the optical elements (42, 52, 62, 72) is such that a plurality of optical elements (42, 52, 62, 72) are arranged at regular intervals along the X-axis direction to form a plurality of X-axis-direction columns (horizontal columns), a plurality of optical elements (42, 52, 62, 72) are arranged at regular intervals along the Y-axis direction to form a plurality of Y-axis-direction columns (horizontal columns), and the optical elements (42, 52, 62, 72) are arranged at each vertex of the square as an orthogonal combination of the plurality of X-axis-direction columns and the plurality of Y-axis-direction columns.

And, a plurality of optical elements (42, 52, 62, 72) arranged in a square array are electrically connected in parallel inside a rectangular housing (41, 51, 61, 71) of a surface illuminator (40, 50, 60, 70), so that all of the optical elements (42, 52, 62, 72) emit light simultaneously by power applied from the outside, thereby illuminating the surface.

Next, in step b), power is supplied to the optical elements (42, 52, 62, 72) to illuminate the subject with the surface illuminator (40, 50, 60, 70), and a rectangular image corresponding to the area of the housing (41, 51, 61, 71) is acquired.

At this time, it is preferable that the surface illuminator (40, 50, 60, 70) is rotated at an angle of 45° based on the incident angle (vertical line) of the image generating unit (80) so as to illuminate the surface diagonally toward the subject to be examined, so that the incident angle of the surface illuminator (40, 50, 60, 70) forms an angle of 45° based on the incident angle (vertical line).

And by placing a light meter at each corner position of the test subject area, the light source arrangement is adjusted to achieve uniform illumination on the test subject, and then a rectangular image is obtained as the image generating unit (80) photographs the surface-illuminated test subject.

The next step is step c), which measures the X-axis shading of the image based on the acquired rectangular image.

If no shading occurs in the rectangular image acquired at this time, the surface illuminator (40, 50, 60, 70) is used as is without spacing correction of the optical elements (42, 52, 62, 72), but if shading occurs in the rectangular image acquired, the length of the shading in the X-axis direction is measured based on the X-axis direction among the shadings in the image acquired by shooting with the image generating unit (80).

Next, in step d), based on the measured X-axis shading, among the plurality of photodiodes (42, 52, 62, 72) arranged in a square shape, the photodiodes (42, 52, 62, 72) arranged in the Y-axis direction are each grouped into one Y-axis direction array, and then the grouped Y-axis direction arrays are moved to adjust the X-axis distance between the photodiodes (42, 52, 62, 72).

The next step is step e), which measures the Y-axis shading of the image based on the acquired rectangular image.

If no shading occurs in the image of the rectangle acquired at this time, the surface illuminator (40, 50, 60, 70) is used as is without spacing correction of the optical elements (42, 52, 62, 72), but if shading occurs in the image of the rectangle acquired, the length of the Y-axis direction of the shading is measured based on the Y-axis direction among the shadings of the image acquired by shooting with the image generating unit (80).

Next, in step f), based on the measured Y-axis shading, among the plurality of photodiodes (42, 52, 62, 72) arranged in a square shape, the photodiodes (42, 52, 62, 72) arranged in the X-axis direction are each grouped into one X-axis direction array, and then the grouped X-axis direction arrays are moved to adjust the Y-axis distance between the photodiodes (42, 52, 62, 72).

Here, an embodiment of the step d) according to the present invention, which is a step of adjusting the X-axis spacing of the optical elements (42, 52, 62, 72), and the step f) according to the present invention, which is a step of adjusting the Y-axis spacing of the optical elements (42, 52, 62, 72), will be examined with reference to FIGS. 3 to 8 as follows.

The first adjustment bar (101), the first adjustment block (102), the first rail (103), the first ruler (104), the second adjustment bar (201), the second adjustment block (202), the second rail (203), and the second ruler (204) can be used to adjust the X-axis and Y-axis spacing of the optical elements (42, 52, 62, 72). In the step d), based on the measured X-axis shading, among the plurality of optical elements (42, 52, 62, 72) arranged in a square shape, the optical elements (42, 52, 62, 72) arranged in the Y-axis direction are each grouped into one Y-axis array, and then the grouped Y-axis arrays are moved to adjust the X-axis spacing of the optical elements (42, 52, 62, 72). A plurality of Y-axis arrays are formed by grouping the optical elements (42, 52, 62, 72) along the X-axis direction rows into one Y-axis direction array in the Y-axis direction. When grouping the optical elements (42, 52, 62, 72) into a Y-axis direction array, they are grouped into a Y-axis direction array by a pair of first adjustment bars (101) and first adjustment blocks (102).

The above-mentioned first adjustment bars (101) are arranged in the Y-axis direction with the optical elements (42, 52, 62, 72) arranged in the Y-axis direction among the plurality of optical elements (42, 52, 62, 72) arranged in a square shape in the center, and are respectively arranged adjacent to the side of the optical elements (42, 52, 62, 72) arranged in the Y-axis direction on the X-axis line, and first adjustment blocks (102) are respectively coupled to both ends of the above-mentioned first adjustment bars (101) on the Y-axis line.

At this time, the pair of first adjustment bars (101) are spaced apart from each other by a length corresponding to the width or depth of the optical elements (42, 52, 62, 72) and are parallel to each other, and the optical elements (42, 52, 62, 72) located in the middle of the pair of first adjustment bars (101) are held in the Y-axis direction by the pair of first adjustment bars (101) and the first adjustment block (102) to form one Y-axis direction array.

Accordingly, a plurality of Y-axis arrays are configured by bundling the optical elements (42, 52, 62, 72) arranged in the Y-axis direction along the X-axis direction row into Y-axis arrays using the above pair of first adjustment bars (101) and first adjustment blocks (102).

And the first adjustment blocks (102) are provided on a first rail (103) having a length in the X-axis direction from the outside of the housing (41, 51, 61, 71), and as the first adjustment block (102) moves along the first rail (103) in the X-axis direction, the optical elements (42, 52, 62, 72) located in the center of the first adjustment bar (101) can move in the X-axis direction while being bundled in a Y-axis array, thereby adjusting the X-axis spacing of the Y-axis arrays.

In addition, a first ruler (104) is provided on the outside of the first rail (103), so that the first adjustment block (102) moving along the first rail can be moved with precise and detailed dimensions, thereby allowing the X-axis spacing of the Y-axis arrays to be precisely adjusted.

At this time, since the X-axis direction length of the shadow shown in the image is proportional to the X-axis direction length of the surface illuminator, it is desirable to arrange the intervals of the Y-axis direction arrays bundled in the Y-axis direction so that the intervals become narrower from one side to the other, that is, toward the shade side.

In addition, in step f), based on the measured Y-axis shading, among the plurality of optical elements (42, 52, 62, 72) arranged in a square arrangement, the optical elements (42, 52, 62, 72) arranged in the X-axis direction are each grouped into one X-axis direction array, and then in the step of moving the grouped X-axis direction arrays to adjust the Y-axis direction spacing of the optical elements (42, 52, 62, 72), the plurality of optical elements (42, 52, 62, 72) arranged in a square arrangement are each grouped into one X-axis direction array along the Y-axis direction columns in the X-axis direction to form a plurality of X-axis direction arrays. When the optical elements (42, 52, 62, 72) are grouped into an X-axis direction array, they are grouped into an X-axis direction array by a pair of second adjustment bars (201) and a second adjustment block (202).

The above-mentioned second adjustment bars (201) are arranged in the X-axis direction, with the optical elements (42, 52, 62, 72) arranged in the X-axis direction among the plurality of optical elements (42, 52, 62, 72) arranged in a square shape in the center, and are respectively arranged adjacent to the Y-axis line side of the optical elements (42, 52, 62, 72) arranged in the X-axis direction, and second adjustment blocks (202) are respectively coupled to both ends of the above-mentioned second adjustment bars (201) on the X-axis line.

At this time, the pair of second adjustment bars (201) are spaced apart from each other by a length corresponding to the width or depth of the optical elements (42, 52, 62, 72) and are parallel to each other, and the optical elements (42, 52, 62, 72) located in the middle of the pair of second adjustment bars (201) are held in the X-axis direction by the pair of second adjustment bars (201) and the second adjustment block (202) to form one X-axis direction array.

Accordingly, a plurality of X-axis-direction arrays are configured by bundling the optical elements (42, 52, 62, 72) arranged in the X-axis direction along the Y-axis direction into X-axis-direction arrays using the above pair of second adjustment bars (201) and second adjustment blocks (202).

And the second adjustment blocks (202) are provided on a second rail (203) having a length in the Y-axis direction on the outside of the housing (41, 51, 61, 71), and as the second adjustment block (202) moves along the second rail (203) in the Y-axis direction, the optical elements (42, 52, 62, 72) located in the center of the second adjustment bar (201) can move in the Y-axis direction while being bundled into an X-axis array, thereby adjusting the Y-axis spacing of the X-axis arrays.

In addition, a second ruler (204) is provided on the outside of the second rail (203), so that the second adjustment block (202) moving along the second rail can be moved with precise and detailed dimensions, thereby allowing the Y-axis spacing of the X-axis arrays to be precisely adjusted.

At this time, since the Y-axis length of the shadow shown in the image is proportional to the Y-axis length of the surface illuminator, it is desirable to arrange the intervals of the X-axis arrays bundled in the X-axis direction so that the intervals become narrower from one side to the other, that is, toward the shade side.

The above-described embodiment is only one example for controlling the spacing between multiple optical elements (42, 52, 62, 72) and is not limited thereto.

Therefore, the method for arranging optical elements equipped in the surface illuminator of the surface defect inspection device according to the present invention arranges a plurality of optical elements (42, 52, 62, 72) in a square array inside the housing (41, 51, 61, 71) of the surface illuminator through the above-described process, and adjusts the spacing between the optical elements based on the shadows appearing in the image of the subject being inspected, so that images for surface defects of various types, shapes, and directions can be significantly distinguished as the shadows are corrected, thereby greatly improving the detection reliability for various surface defects of the subject being inspected.

Therefore, the surface defect inspection device (1) according to one embodiment of the present invention is equipped with a total of four surface illuminators (first to fourth surface illuminators), and when implementing the optical element arrangement method equipped in the surface illuminators of the surface defect inspection device, it is preferable to adjust the optical element arrangement of the first surface illuminator (40) through the above-described series of processes, and then sequentially adjust the optical element arrangement of the second surface illuminator (50), the third surface illuminator (60), and the fourth surface illuminator (70).

In addition, even if the illumination is distributed uniformly in the inspection target area, the uniformity of brightness and incident angle in the sample image is reduced due to the Gaussian and Lambertian effects of the light incident on the camera lens (optical system).

A sample image correction method of a surface defect inspection device according to one embodiment of the present invention aims to resolve brightness unevenness and incident angle unevenness caused by the Gaussian and Lambertian effects.

A method for correcting a sample image of a surface defect inspection device according to an embodiment of the present invention with reference to FIGS. 10 to 15 relates to a method for correcting a sample image of a surface defect inspection device that inspects a surface defect of a subject by photographing the subject while illuminating the subject with light and analyzing the photographed sample images, the method comprising the steps of: i) photographing the subject multiple times while sequentially incident light from different directions based on an incident area of the light among the subject, thereby obtaining sample images each corresponding to a different incident angle of the light; ii) optimizing the brightness of each of the obtained sample images; iii) optimizing the incident angle of the brightness-optimized sample images; and iv) obtaining a corrected final sample image.

First, in step i), the test subject is photographed multiple times while sequentially incident light from different directions based on the light incidence area among the test subject, thereby obtaining sample images corresponding to different light incidence angles.

At this time, it is desirable to photograph the subject while sequentially incident light from each of the first, second, third, and fourth quadrants of the incident area based on the incident area of the light among the subject's images, and when acquiring a total of four sample images, it is desirable to collect information on the incident angle and brightness of the light together.

The next step is step ⅱ, where the brightness of each acquired sample image is optimized.

At this time, the entire area of the acquired sample image is divided into pixel units so that the same brightness is displayed throughout the entire area, and brightness weights are applied to the corresponding pixels to optimize the brightness of the acquired sample image.

Referring to Fig. 11, the process of optimizing the brightness of acquired sample images is examined. First, in step ⅱ-1), the average pixel value of the pixels forming the sample image is extracted.

In other words, when the acquired sample image is divided into pixel units of the corresponding size, it is converted into an image whose gray level can be determined, and each divided pixel is assigned a gray level corresponding to the brightness value, and the average pixel value of the average brightness is extracted based on the gray level assigned to each pixel.

The next step is step ⅱ-2), which searches for the relatively brightest pixel among the pixels forming the sample image and extracts the max pixel value based on the gray level of the searched pixel.

At this time, each segmented pixel is assigned a gray level corresponding to the brightness value, and based on the gray level assigned to each pixel, the pixel with the highest maximum brightness with the relatively highest gray level is searched for, and the max pixel value is extracted with the gray level of the searched pixel.

The next step is step ⅱ-3), which searches for the relatively darkest pixel among the pixels forming the sample image and extracts the min pixel value based on the gray level of the searched pixel.

At this time, each segmented pixel is assigned a gray level corresponding to its brightness value, and based on the gray level assigned to each pixel, the pixel with the lowest minimum brightness with a relatively low gray level is searched for, and the min pixel value is extracted with the gray level of the searched pixel.

The next step is step ⅱ-4), which calculates the Wmax pixel value, which is a weight for the extracted max pixel value to become the average pixel value, and calculates the Wmin pixel value, which is a weight for the extracted min pixel value to become the average pixel value.

At this time, the average pixel values extracted in steps ⅱ-1), ⅱ-2), and ⅱ-3) are used to calculate the Wmax pixel value for adjusting the max pixel value to the average pixel value.

And, using the average pixel values extracted in steps ii-1), ii-2), and ii-3), respectively, the Wmin pixel value for adjusting the min pixel value to the average pixel value is calculated.

The next step is step ⅱ-5), which calculates the distance between the darkest pixel (min pixel value) and the brightest pixel (max pixel value).

Here, the relatively longest distance is Dmax, and the relatively shortest distance is Dmin, which is 0.

The following is ⅱ-6), which calculates the change in weight per unit pixel distance (Kstep).

At this time, the change in weight per unit pixel distance (Kstep) is calculated using the following [Mathematical Formula 1].

Here, Kstep is the change in weight per unit pixel distance, Wmax is the weight for the max pixel value to become the average pixel value, Wmax is the weight for the max pixel value to become the average pixel value, and Dmax is the distance between the brightest pixel (max pixel value) and the darkest pixel (min pixel value).

The next step is step ⅱ-7), which calculates the weight (K) of each pixel that makes up the sample image.

At this time, the weight (K) of each pixel is calculated using the following [Mathematical Formula 2].

Here, K is the weight of each pixel, Wmin is the weight for the min pixel value to become the average pixel value, Kstep is the change in weight per unit pixel distance, and Distance is the distance between each pixel based on the brightest pixel.

A map for brightness weights can be generated by the above process.

The next step is step ⅱ-8), where the calculated weights are applied to each pixel forming the sample image to optimize the brightness of the sample image.

Here, it is desirable that the above brightness be optimized for each sample image corresponding to different light incident angles.

The next step is step ⅲ, which involves angular optimization of brightness-optimized sample images.

At this time, referring to Fig. 12, at least three sample images are selected from among the sample images for which brightness optimization has been performed, and the incident angle is optimized using the selected sample images for which brightness optimization has been performed.

Looking at the above-mentioned incident angle optimization process, first, ⅲ-1) prepare at least 3 sample images out of a total of 4 sample images with different incident angles of light.

The sample images prepared at this time are sample images that have been optimized for brightness.

And then, in step ⅲ-2), the incident angle of light reflected in each sample image is calculated.

At this time, referring to Fig. 15, the incident angles of light at the three vertices closest to the incident direction of light are measured for each sample image, and since the angular tendency can be known within an image, the incident angle can be predicted for each pixel of the image.

The next step is step ⅲ-3), where the normal vector of the corresponding pixel is calculated using photomatrice stereo modeling based on the actual incident angle obtained from three sample images for each pixel.

Here, p is the albedo constant, N is the normal vector, S is the inverse matrix of the light 3-way matrix, and I is the brightness values of each pixel of the three sample images.

Next, in step ⅲ-4), the normal vector of the obtained pixel is inputted using photomatrice stereo modeling, and the incident angles of light set by the user to the illuminator are inputted to obtain a sample image with a new light brightness value (gray level), i.e., an incident angle-corrected sample image.

If we merge the linear equations, we get [Mathematical Equation 5] below.

In other words, a uniform sample image can be obtained as an incident angle of the surface illuminator as in Fig. 14.

The next step is step ⅳ), where the corrected final sample image is obtained.

Therefore, the sample image correction method of the surface defect inspection device that performs the above process corrects the sample image by optimizing the brightness and optimizing the incident angle of the Gaussian and Lambertian effects (configurable as a lookup table), thereby eliminating the brightness unevenness and incident angle unevenness of the sample image due to the Gaussian and Lambertian effects.

Although the present invention has been described with reference to the embodiments illustrated in the drawings, these are merely exemplary, and those skilled in the art will understand that various modifications and equivalent other embodiments are possible therefrom. Therefore, the true technical protection scope of the present invention should be determined by the technical idea of the appended claims.

1: Surface defect inspection device 10: Frame part
20: Optical module support 30: Transport section
31: Transfer shaft 32: Transfer plate
33: Transfer motor 34: Connecting plate
40, 50, 60, 70: Surface illuminator 41, 51, 61, 71: Housing
42, 52, 62, 72: Photoelectric element 80: Image generation unit
101: 1st adjustment bar 102: 1st adjustment block
103: 1st rail 104: 1st ruler
201: 2nd adjustment bar 202: 2nd adjustment block
203: 2nd rail 204: 2nd ruler

Claims (11)

In a method for arranging optical elements provided in a surface illuminator of a surface defect inspection device of a subject,
a) A step of arranging a plurality of optical elements in a rectangular array arbitrarily inside a rectangular housing of a surface illuminator;
b) A step of applying power to the above optical elements to illuminate the subject with a surface illuminator and obtaining a rectangular image corresponding to the housing area;
c) A step of measuring the X-axis direction shading of the image based on the acquired rectangular image;
d) a step of grouping the Y-axis-arranged optical elements among the plurality of rectangularly arranged optical elements into one Y-axis-array based on the measured X-axis shading, and then moving the grouped Y-axis-arrays to adjust the X-axis-direction spacing of the optical elements;
e) a step of measuring the Y-axis direction shading of the image based on the acquired rectangular image; and
f) A method for arranging optical elements in a surface illuminator of a surface defect inspection device, comprising: a step of grouping optical elements in an X-axis array among a plurality of rectangularly arranged optical elements into one X-axis array based on the measured Y-axis shading; and then moving the grouped X-axis arrays to adjust the Y-axis spacing of the optical elements.
In claim 1,
In the step of measuring the X-axis direction shading of the image based on the acquired rectangular image in the above step c),
Measure the length of the shade in the X-axis direction based on the X-axis direction of the image above,
In the step d) above, based on the measured X-axis shading, among the plurality of rectangularly arranged optical elements, the optical elements in the Y-axis array are each grouped into one Y-axis array, and then the grouped Y-axis arrays are moved to adjust the X-axis spacing of the optical elements.
A method for arranging optical elements in a surface illuminator of a surface defect inspection device, wherein the intervals between Y-axis arrays, which are tied in the Y-axis direction, become narrower as they go toward the shaded side.
In claim 1,
In the step of measuring the Y-axis direction shading of the image based on the acquired rectangular image in the above step e),
Measure the length of the shade in the Y-axis direction based on the Y-axis direction of the image above,
In the step f) above, based on the measured Y-axis shading, among the plurality of rectangularly arranged optical elements, the optical elements in the X-axis array are each grouped into one X-axis array, and then the grouped X-axis arrays are moved to adjust the Y-axis spacing of the optical elements.
A method for arranging optical elements in a surface illuminator of a surface defect inspection device, wherein the intervals between X-axis arrays, which are grouped in the X-axis direction, become narrower as they go toward the shaded side.
In claim 1,
b) In the step of applying power to the optical elements and illuminating the subject with a surface illuminator, a rectangular image corresponding to the housing area is obtained.
A method for arranging optical elements in a surface illuminator of a surface defect inspection device, which measures the illuminance of an incident area by arranging an illuminance meter at each corner position of the surface of the subject to be inspected, adjusts a light source to achieve uniform illuminance over the entire surface of the subject to be inspected, and then acquires an image of the surface-illuminated subject to be inspected.
A method for correcting a sample image of a surface defect inspection device that inspects surface defects of a subject by illuminating the subject with light using any one of the optical element array methods of claims 1 to 4, photographing the subject, and analyzing the photographed sample images,
i) A step of taking multiple pictures of the subject while sequentially incident light from different directions based on the light incidence area among the subject, and obtaining sample images corresponding to different light incidence angles;
ⅱ) A step of optimizing the brightness of each acquired sample image;
ⅲ) A step of optimizing the angle of incidence of each brightness-optimized sample image; and
ⅳ) A method for correcting a sample image of a surface defect inspection device, comprising the step of obtaining a corrected final sample image.
In claim 5,
In the step ⅱ above, the step of optimizing the brightness of the acquired sample image is as follows:
ⅱ-1) A step of extracting an average pixel value from among the pixels forming a sample image;
ⅱ-2) A step of searching for a relatively brightest pixel among the pixels forming a sample image and extracting the max pixel value based on the gray level of the searched pixel;
ⅱ-3) A step of searching for a relatively darkest pixel among the pixels forming a sample image and extracting a min pixel value based on the gray level of the searched pixel;
ⅱ-4) A step of calculating the Wmax pixel value, which is a weight for adjusting the extracted max pixel value to the average pixel value, and calculating the Wmin pixel value, which is a weight for adjusting the extracted min pixel value to the average pixel value;
ⅱ-5) A step of calculating the distance between the darkest pixels (min pixel value) based on the brightest pixel (max pixel value);
ⅱ-6) A step for calculating the change in weight per unit pixel distance (Kstep);
ⅱ-7) A step of calculating the weight (K) of each pixel forming the sample image;
ⅱ-8) A method for correcting a sample image of a surface defect inspection device, comprising: a step of applying the calculated weight to each pixel forming the sample image to optimize the brightness of the sample image.
In claim 6,
The change in weight per unit pixel distance (Kstep) is a sample image correction method of a surface defect inspection device calculated by [Mathematical Formula 1] below.
[Mathematical Formula 1]
Kstep=(Wmin-Wmax)/Dmax
(Here, Kstep is the change in weight per unit pixel distance, Wmin is the weight for the min pixel value to become the average pixel value, Wmax is the weight for the max pixel value to become the average pixel value, and Dmax is the distance between the darkest pixels based on the brightest pixel.)
In claim 7,
The weight (K) of each pixel above is a sample image correction method of a surface defect inspection device calculated by [Mathematical Formula 2] below.
[Mathematical formula 2]
K= Wmax+Kstep×Distance
(Here, K is the weight of each pixel, Wmax is the weight for the max pixel value to become the average pixel value, Kstep is the change in weight per unit pixel distance, and Distance is the distance between each pixel based on the brightest pixel.)
In claim 8,
In the step iii) above, the angle of incidence of the sample images for which brightness optimization has been performed is optimized.
ⅲ-1) A step of preparing at least three sample images having different light incidence angles among the sample images for which brightness optimization has been performed;
ⅲ-2) A step of measuring the actual incident angles of light of each sample image and confirming the incident angle for each pixel of the sample images;
ⅲ-3) A step of calculating a pixel normal vector using three or more gray levels and actual angles of the pixel based on the actual incidence angle obtained for each pixel using photomatrice stereo modeling;
ⅲ-4) A method for correcting a sample image of a surface defect inspection device, comprising: a step of obtaining a sample image in which the incident angle is optimized with a new light brightness by inputting a pixel normal vector and arbitrarily set incident angles of light using a photomatrice stereo model;
In claim 9,
Based on the actual angle of incidence obtained for each pixel in step ⅲ-3 above, in the step of calculating the pixel normal vector using photomatrice stereo modeling with three or more gray levels and actual angles of the corresponding pixel,
The pixel normal vector is a sample image correction method of a surface defect inspection device calculated by [Mathematical Formula 3] below.
[Mathematical Formula 3]

(Here, ρ is the albedo constant, N is the normal vector, S is the inverse matrix of the light 3-way matrix, and I is the brightness values of each pixel of the three sample images.)
In claim 9,
In the step ⅲ-4) above, using the photomatrice stereo model, a sample image optimized for the incident angle with a new light brightness is obtained by inputting the pixel normal vector and the arbitrarily set incident angles of light.
A sample image correction method of a surface defect inspection device in which the incident angle is corrected by [Mathematical Formula 4] below.
[Mathematical formula 4]

,

,

,
(Here, ρ is the albedo constant, N is the normal vector, S is the inverse matrix of each light 3-way matrix of the three sample images, and I is the brightness values of each pixel of the three sample images.)

Images