KR100666468B1 - Method and apparatus for inspecting pattern defect of transparent plate having periodic scattering patterns - Google Patents

Abstract

A testing method and a testing apparatus are provided to differentiate a real defect in a transparent plate having periodical diffusion patterns from a similar defect caused by a surface particle or the abnormality of a passivation layer. A transmissive lighting image of a transparent plate is obtained using a transmissive lighting unit(S110). A lateral lighting image of the transparent plate is obtained using a lateral transmissive lighting unit(S120). A first defect candidate, which is not matched to the periodicity of patterns, is extracted from the transmissive lighting image(S130). A second defect candidate, which is not matched to the periodicity of the patterns, is extracted from the lateral lighting image(S140). A real defect is differentiated from a similar defect through comparison between the first defect and the second defect, thereby detecting the defect of the patterns(S150).

Description

Method and apparatus for inspecting pattern defect of transparent plate having periodic scattering patterns

1 is an exploded perspective view showing an example of a backlight unit.

2 is a flowchart of a test method according to the present invention.

3 is a view showing an inspection apparatus according to the present invention.

4A shows the light path of transmitted illumination.

4B is a diagram illustrating an example of a pattern image acquired by transmission illumination.

5a shows the light path of the side lighting.

6B is a diagram illustrating an example of a pattern image acquired by side lighting.

Fig. 6 is a diagram showing a relationship between processing in an actual image and processing in frequency.

7 is a view showing a result of applying a spatial filter to the side illumination image according to an embodiment of the present invention.

8A and 8B are diagrams showing examples of a transmission illumination image and a side illumination image of a defect of a defect respectively.

9A and 9B are diagrams showing examples of a transmission illumination image and a side illumination image of air gap defects, respectively.

10A and 10B are diagrams showing examples of a transmission illumination image and a side illumination image of pattern printing surface dust, respectively.

11A and 11B are diagrams showing examples of a transmission illumination image and a side illumination image of dust on the opposite side of the pattern printing, respectively.

FIG. 12 is a detailed flowchart of a comprehensive determination step of the inspection method illustrated in FIG. 2 and illustrates a conditional expression for distinguishing a pattern defect from other similar defects in an image processing unit of the pattern defect inspection apparatus according to the present invention.

 <Explanation of symbols for the main parts of the drawings>

200 ... pattern defect inspection device 205 ... scattering pattern

210 ... transparent plate 220 ... transparent illuminator

230 ... side lighting 235 ... reflector

240 ... Image Acquisition Unit 250 ... Image Processing Unit

The present invention relates to an inspection method and apparatus for detecting a pattern defect of a transparent plate having a periodic scattering pattern, and more particularly, a light guide plate of a backlight unit (BLU) for a flat panel display. (Light Guide Plate: LGP) relates to an optical inspection method and apparatus for detecting a pattern defect.

If the shape and size of the pattern is different from the design, it is called a pattern defect. In most optical inspection apparatuses that automatically inspect such pattern defects, foreign materials such as dust on the product are often detected as pattern defects along with the actual pattern defects.

Take the case of the light guide plate, which is a main component of the backlight unit.

In general, a light-receiving flat panel display or a light box such as an LCD injects a liquid crystal having an intermediate property between a solid and a liquid between two thin glass plates, thereby changing the arrangement of liquid crystal molecules during power supply. Unlike PDP, FED, organic EL, etc., it is a non-light emitting type (light receiving element) and thus cannot be used without a separate light emitting device. There is a need for a backlight unit in the form of a sustainable surface light source.

A key component of the general backlight unit U shown in FIG. 1 is light emitting means L for determining luminance and chromaticity, such as a cold cathode fluorescent lamp (CCFL) or an LED, and the light emitting means ( L) reflector 20 to increase the efficiency of the light generated in the L), and the linear light received from the reflector 20 is specially designed on the back surface of the transparent acrylic plate of PMMA resin to form a surface light source form the entire image display surface The dot pattern 5 is the light guide plate 10 processed. In addition, the surface light source generated by the three elements, that is, the light emitting means L, the reflector 20, and the light guide plate 10, efficiently diffuses light by the diffusion plate 30 provided on the light guide plate 10. do. In addition, an upper surface of the diffusion plate 30 may be provided with a separate prism plate (not shown) to increase the light collecting efficiency. The method of forming the dot pattern 5 on the light guide plate 10 may include a method of forming ink into a predetermined pattern by a printing method, a method of melting a light guide plate raw material resin and injecting the mold into a mold having a predetermined pattern, and forming a laser. And a method of processing a furnace pattern.

Important components of the backlight unit, such as the light emitting means L, the reflector 20, and the light guide plate 10, are delivered with a protective film for protection of raw materials so that the protective film is removed immediately before assembly of the backlight unit. Manufacturers with light guide plate pattern printing technology supply acrylic transparent discs with a protective film on the top and bottom to remove the protective film on one side, print a periodic scattering pattern on the surface, and then apply a protective film on it to assemble the backlight unit. Transfer to process. Pattern defects are detected by inspecting the light guide plate's pattern between printing and conveyance. Existing optical inspection devices do not distinguish between actual pattern defects and dust, and thus scratches and stains that may appear on foreign matters and protective films on the light guide plate. The defects are classified as defects and processed for defects. By the way, foreign matters attached to the upper and lower light guide plates, scratches and stains on the protective film (hereinafter, similar defects) are sufficiently removed by the process of removing the protective film and cleaning the light guide plate in the backlight unit assembly process. For this reason, conventionally, there is a problem in that product quality which is not distinguished from actual pattern defects and similar defects is not disposed of, and thus the good product is discarded and the productivity of the product is excessively reduced.

In order to solve this problem, there is an example in which a pattern defect and a foreign material are distinguished using transmitted light and reflected light. However, in the case of the light guide plate, the pattern printing surface causes diffuse reflection of the reflected light, thereby distorting the pattern information. In addition, if the condition is to be inspected while the protective film is attached to the opposite surface of the light guide plate, bubbles that may be formed between the light guide plate and the protective film, scratches and stains of the protective film make it difficult to detect accurate pattern defects.

For this reason, at present, an operator is put in after the light guide plate pattern printing process to visually inspect whether there is an abnormality in the printed pattern. In the light guide plate for the 19-inch backlight unit, the number of dot patterns is about 50,000, and it is difficult to quantitatively control the quality while inspecting them.

The technical problem to be achieved by the present invention is to provide a pattern defect inspection method that can distinguish pattern defects of a transparent plate having a periodic scattering pattern, such as a light guide plate, with or without a protective film, from similar defects caused by surface foreign matter or protective film abnormalities. .

Another object of the present invention is to provide a pattern defect inspection apparatus capable of realizing such a pattern defect inspection method.

The pattern defect inspection method according to the present invention for achieving the above technical problem, in the pattern defect inspection method of a transparent plate having a periodic scattering pattern, (a) using a transmission illumination to obtain a transmission illumination image of the transparent plate step; (b) acquiring side lighting images of the transparent plate using side lighting; (c) extracting a first defect candidate that deviates from the periodicity of a pattern in the transmitted illumination image; (d) extracting a second defect candidate from the side illumination image that is inconsistent with the periodicity of the pattern; And (e) detecting pattern defects by distinguishing between actual pattern defects and similar defects provided by an inspection environment by comparing the first defect candidates with the second defect candidates.

In the inspection method according to the present invention, in the steps (c) and (d), a spatial filter suitable for extracting the nonuniformity of the pattern through the difference in brightness of the surrounding pixels spaced a predetermined period is applied, and the spatial filter is imaged. When convolution with the data and the resulting filtering data exceeds a specific threshold, it is preferable to extract a region corresponding to the image data as a defect candidate.

In the inspection method according to the present invention, the defect candidates brighter than the surrounding pixels spaced apart from the predetermined period are classified as white defects, and the candidates darker than the surrounding pixels spaced apart from the predetermined period are classified as black defects. Classify as In the step (e), among the intersections of the first defect candidate and the second defect candidate, a defect candidate classified as a black defect in the transmission illumination image is classified as a white defect in the side illumination image, or the transmission Only when a defect candidate classified as a white defect in the illumination image is classified as a black defect in the side illumination image, the defect candidate is determined as a pattern defect.

The pattern defect inspection apparatus according to the present invention for achieving the another technical problem, in the pattern defect inspection apparatus of the transparent plate having a periodic scattering pattern, the transmission to irradiate light from the lower portion of the transparent plate toward the transparent plate Lighting unit; Side lighting units for irradiating light toward the transparent plate from both sides or one side of the transparent plate; Acquire a side illumination image of the transparent plate using light transmitted from the transmissive lighting unit and transmitted through the transparent plate by using light transmitted from the side lighting unit and scattered and reflected inside the transparent plate. An image acquisition unit; And extracting a first defect candidate that deviates from the periodicity of the pattern in the transmitted illumination image, extracting a second defect candidate deviating from the periodicity of the pattern in the side illumination image, and comparing the first defect candidate and the second defect candidate. It includes an image processor for detecting a pattern defect by distinguishing between the actual pattern defect and similar defects provided by the inspection environment.

In the inspection apparatus according to the present invention, the image processor includes a spatial filter suitable for extracting the nonuniformity of the pattern through the difference in the brightness of the surrounding pixels at a predetermined period, and convoluting the spatial filter with the image data as a result If the filtering data, which is a value, exceeds a specific threshold, a portion corresponding to the image data is extracted as a defect candidate. In this case, it is preferable that the spatial filter is designed to emphasize high frequency components caused by pattern defects among image data converted to spatial frequencies based on periodicity of regular patterns.

In the inspection apparatus according to the present invention, the image processing unit preferably extracts the second defect candidate at the position of the first defect candidate. A reflective plate may be further included below the transparent plate to collect scattered and reflected light from the inside of the transparent plate to the upper portion of the transparent plate. The light source of the transmissive lighting unit and the side lighting unit may have a wavelength in the visible light region.

Pattern defect inspection method and apparatus according to the present invention can be applied regardless of the protective film of the transparent plate. Accordingly, the transparent plate may or may not have a protective film attached to the upper or lower portion or the upper and lower portions thereof.

Advantages and features of the present invention, and methods for achieving them will be apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various forms, and only the present embodiments are intended to complete the disclosure of the present invention and to those skilled in the art to fully understand the scope of the invention. It is provided to inform you.

Hereinafter, the pattern defect inspection method and apparatus according to the present invention will be described first with reference to FIGS. 2 and 3, and then the details of the pattern defect inspection method and apparatus will be described in detail.

2 is a flowchart of a pattern defect inspection method according to the present invention.

As shown in FIG. 2, the pattern defect inspection method according to the present invention is a pattern defect inspection method of a transparent plate having a periodic scattering pattern, and includes a transmission illumination image acquisition step s110 and a side illumination image acquisition step s120. The method may include extracting the first defect candidate from the transmitted illumination image (S130), extracting the second defect candidate from the side illumination image (S140), and detecting the pattern defect by the comprehensive determination (S150).

In the transmission illumination image acquisition step (s110), the transmission illumination image of the transparent plate is acquired using the transmission illumination. Preferably, light is irradiated from the lower portion of the transparent plate toward the transparent plate, and light transmitted through the transparent plate is obtained from the upper portion of the transparent plate to obtain a transmission illumination image.

In the side illumination image acquisition step (s120), the side illumination image of the transparent plate is acquired using side illumination. Preferably, light is irradiated toward both sides or one side of the transparent plate toward the transparent plate, and light scattered and reflected inside the transparent plate is obtained from the upper side of the transparent plate. At this time, in order to collect the light scattered and reflected inside the transparent plate to the upper portion of the transparent plate may be provided with a reflecting plate below the transparent plate.

In the extracting of the first defect candidate (S130), a defect candidate that is inconsistent with the periodicity of the pattern is first extracted from the transmitted illumination image. In the extracting of the second defect candidate (S140), the defect candidate that is inconsistent with the periodicity of the pattern is secondarily extracted from the side illumination image. At this time, the step of extracting the second defect candidate from the position of the first defect candidate obtained in the first defect candidate extracting step s130 (s140) may save inspection time. In the detecting of the pattern defect by the comprehensive determination (S150), the pattern defect is detected by distinguishing the actual pattern defect from the similar defect provided by the inspection environment by comparing the first defect candidate and the second defect candidate.

Such a pattern defect inspection method can be realized by using the pattern defect inspection apparatus shown in FIG. 3 is a diagram illustrating a configuration of a pattern defect inspection apparatus according to the present invention. The pattern defect inspection method and apparatus according to the present invention can be applied with or without a protective film of the transparent plate. Therefore, the transparent plate to be inspected may or may not have a protective film attached to the upper or lower portion or the upper and lower portions thereof.

Referring to FIG. 3, the pattern defect inspection apparatus 200 according to the present invention is a pattern defect inspection apparatus of the transparent plate 210 having a periodic scattering pattern 205, and includes a transmission illumination unit 220 and a side illumination unit 230. ), An image acquisition unit 240, and an image processing unit 250. The light sources of the transmissive lighting unit 220 and the side lighting unit 230 may have a wavelength in the visible region.

The transmissive lighting unit 220 irradiates light from the lower portion of the transparent plate 210 toward the transparent plate 210, and the image acquisition unit 240 is irradiated from the transmissive lighting unit 220 to transmit the light through the transparent plate 210. Using to obtain a transmission illumination image of the transparent plate 210. Therefore, the transmission illumination image acquisition step (step s110 of FIG. 2) of the pattern defect inspection method according to the present invention may be performed using the transmission illumination unit 220 and the image acquisition unit 240.

The side lighting unit 230 irradiates light toward the transparent plate 210 from both sides of the transparent plate 210 (the example shown may be related to both sides or may be installed on one side), and the image acquisition unit 240 ) Also obtains a side illumination image of the transparent plate 210 by using light reflected from the side lighting unit 230 and scattered and reflected inside the transparent plate 210. Therefore, the side illumination image acquisition step (step s120 of FIG. 2) of the pattern defect inspection method according to the present invention may be performed using the side illumination unit 230 and the image acquisition unit 240. A reflective plate (see 235 of FIG. 5A) may be further included under the transparent plate 210 to collect light scattered and reflected from the inside of the transparent plate 210 to the upper portion of the transparent plate 210.

The image processor 250 extracts a first defect candidate that deviates from the periodicity of the pattern in the transmitted illumination image, extracts a second defect candidate that deviates from the periodicity of the pattern in the side illumination image, and extracts a first defect candidate of the first defect candidate and the second defect candidate. The comparison detects pattern defects by distinguishing between actual pattern defects and similar defects provided by the inspection environment. Therefore, the first defect candidate extraction step, the second defect candidate extraction step, and the comprehensive determination step (steps s130 to s150 of FIG. 2) of the pattern defect inspection method according to the present invention may be performed using the image processor 250. .

Now, the details of the pattern defect inspection method and apparatus according to the present invention will be described in detail.

Given a transparent plate 210 having a periodic scattering pattern 205 as shown in FIG. 3, in the transmission illumination image acquisition step (step s110 of FIG. 2), the transparent illumination unit 220 positioned below the transparent plate 210 is disposed. Drive to obtain a transmitted illumination image. As shown in FIG. 4A, when light is irradiated from the lower portion of the transparent plate 210 toward the transparent plate 210, the scattering pattern 205 printed on the transparent plate 210 is irradiated from the transmission illumination unit 220. Block the light. For this reason, as shown in FIG. 4B, the scattering pattern 205 may have a transmission illumination image T having a darker shape than the transparent plate 210. (FIG. 4B shows a dot arrangement of a 'circular-dot-on-honeycomb' type.)

In the side illumination image acquisition step (step s120 of FIG. 2), a side illumination image is obtained by driving the side illumination unit 230 located on the side of the transparent plate 210. As shown in FIG. 5A, the light irradiated toward the transparent plate 210 from the side of the transparent plate 210 hits the scattering pattern 205 and undergoes scattering reflection inside the transparent plate 210. In this case, the reflective plate 235 may be installed below the transparent plate 210 to collect light scattered and reflected inside the transparent plate 210 to the upper portion of the transparent plate 210. Due to the scattering characteristics of the scattering pattern 205 as shown in FIG. 5A, the side illumination image S as shown in FIG. 5B in which the scattering pattern 205 is relatively brighter than the transparent plate 210 is obtained.

In the extracting of the first defect candidate (S130) and the extracting the second defect candidate (S140) of FIG. 2, a spatial filter suitable for extracting the nonuniformity of the pattern is applied through the difference in brightness of the surrounding pixels spaced apart from each other. In addition, it is preferable to extract a region corresponding to the image data as a defect candidate when the spatial filter is convolved with the image data and the resulting filtering data exceeds a specific threshold. Accordingly, a spatial filter suitable for extracting a nonuniformity of a pattern through brightness differences with surrounding pixels spaced apart from a predetermined period is designed and included in the image processing unit 250 of FIG. 3, and the spatial filter is included in the image acquisition unit 240. If the filtering data exceeds a specific threshold by convolving with the acquired image data from the transmitted illumination image T and the side illumination image S, a portion corresponding to the image data is selected as a defect candidate.

As described above, according to an exemplary embodiment of the present invention, a defect candidate is extracted by applying a spatial filter. Details thereof are as follows.

Fig. 6 is a diagram showing a relationship between processing in an actual image and processing in frequency. As shown in FIGS. 6A and 6B, two-dimensional image data (original image) can be converted into a spatial frequency domain. Spatial frequency refers to the degree of change of image brightness (the rate of change of pixel value). High frequency component means a delicate image with a lot of change in brightness, and low frequency means a poor image with little change in brightness.

As is well known, a filter in the frequency domain serves to pass certain components and to block some components, which can be used to improve the original image by manipulating a constant frequency region by filtering the original frequency components.

Convolving the spatial filter as shown in Fig. 6C with the image data has the same effect as the frequency domain filtering as shown in Figs. 6D and 6E. Here, the spatial filter is preferably designed to emphasize (high pass) the high frequency component represented by the pattern defect among the image data converted to the spatial frequency based on the periodicity of the regular pattern.

7 is a view showing a result of applying a spatial filter to the side illumination image according to an embodiment of the present invention. The side illumination image shows a total of three defects, including two defects (described below) and one defect missing the entire pattern shape. The result of applying the defect detection spatial filter is normalized from 0 to (N-1) N scale (e.g., N = 256), and values are concentrated near (N / 2) when the target pattern is regular. If there is a component that is irregularly brighter than the surroundings, it is larger than (N / 2), and the stronger the brighter component, the closer to (N-1). If there is an irregularly darker component than the surroundings, the value is smaller than (N / 2). The stronger the darker component, the closer to zero. The threshold is applied to classify the image as a white defect if it has a bright value compared to the surrounding pixels spaced apart from the periodicity of the pattern image. black defects). Therefore, when using the spatial filter as in the embodiment of the present invention, it is possible to determine whether or not the current image is defective without storing the reference image without defect in advance.

In the first defect candidate extraction step (step s130 of FIG. 2) performed by the image processor 250, the defect candidate is extracted from the transmission illumination image T using the above method, and the second defect candidate extraction step (of FIG. 2). In step s140, a defect candidate is extracted from the side illumination image S using the above method. In particular, as mentioned above, a defect candidate is extracted by defining a test area to a defect candidate position selected in the first defect candidate extraction step (step s130 of FIG. 2) in the second defect candidate extraction step (step s140 of FIG. 2). Internal inspection time can be shortened. In this case, if it is normal in the first defect candidate extraction step (step s130 of FIG. 2), the process may end without the second defect candidate extraction step (step s140 of FIG. 2).

In the comprehensive determination step (step s150 of FIG. 2) of the image processor 250, the defect candidates are extracted as the defect candidates in the first defect candidate extraction step (step s130 of FIG. 2) and the second defect candidate extraction step (step s140 of FIG. 2). It is determined whether the area is an actual pattern defect or a similar defect provided by the inspection environment such as dust or protective film abnormality on the surface.

Among defects occurring beyond the shape of the transparent plate printing pattern, a pattern having a smaller size than a normal pattern is referred to as a defect defect, and a pattern having a larger size than the normal pattern is referred to as an air defect. 8A is an example of the image of the defect defect 501 obtained by the transmission illumination, and FIG. 8B is an example of the image of the defect defect 502 obtained by the side lighting. FIG. 9A is an example of the image of the defect defect 601 obtained by the transmission illumination, and FIG. 9B is an example of the image of the defect defect 602 obtained by the side illumination.

In the transmission illumination image of FIG. 8A, the defect defect 501 is represented as a white defect in which a defect portion in which a pattern is not formed is brighter than pixels of a normal pattern around it. In the transmission illumination image of FIG. 9A, the air gap defect 601 is represented as a black defect having a darker characteristic than surrounding pixels in which a defect portion in which a pattern is formed larger than a normal size is spaced a predetermined period apart. In the side illumination image of FIG. 8B, the defect defect 502 appears as a black defect, and in the side illumination image of FIG. 9B, the air defect 602 appears as a white defect.

Thus, it turns out that the defect which generate | occur | produced more than the shape of a transparent plate printing pattern turns into a black defect, and a black defect turns into a white defect according to an illumination system. In contrast, in the example of dust, the brightness difference with the surrounding pixels is maintained as in FIGS. 10A, 10B, 11A, and 11B.

10A and 10B are diagrams showing examples of a transmission illumination image and a side illumination image of a pattern printing surface dust. 11A and 11B are diagrams showing examples of a transmissive illumination image and a side illumination image of dust on the opposite side of pattern printing. In FIGS. 10A, 10B, 11A and 11B, the lines across the drawing show an example of hair, ie dust, and the triangles in FIGS. 11A and 11B are markers for indicating the location of the hair. .

The black defects 701 and 801 caused by the dust in the transmission illumination images of FIGS. 10A and 11A may be seen as black defects 702 and 802 in the side illumination images of FIGS. 10B and 11B.

When the protective film is attached to the opposite surface of the transparent plate having a scattering pattern, such as a light guide plate, scratches of the protective film or bubbles formed between the transparent plate and the protective film are seen with reference to FIGS. 10A, 10B, 11A, and 11B. As in the case of the present invention, the brightness difference with the neighboring pixels spaced apart is maintained. That is, even if the illumination changes, the characteristics of black defects and white defects are not reversed.

Therefore, in the case of comparing the transmission illumination image and the side illumination image as in the present invention, defects occurring beyond the shape of the transparent plate printing pattern change the characteristics of the defects according to the illumination method, whereas bubbles or scratches of the protective film Similar defects caused by the abnormality of the protective film, it can be seen that the characteristics of the defect does not change according to the illumination method. Accordingly, when both the transmission illumination image and the side illumination image are acquired and compared as in the present invention, the pattern defect and the similar defect due to dust, bubbles, scratches, etc. can be distinguished.

12 shows the defects in the first defect candidate extraction step (step s130 in FIG. 2) and the second defect candidate extraction step (step s140 in FIG. 2) in the comprehensive determination step (step s150 in FIG. 2) of the image processor 250. The conditional expressions used to determine whether the region extracted as a candidate is an actual pattern defect or a similar defect provided by an inspection environment such as dust or protective film abnormality on the surface are shown together.

Referring to FIG. 12, which summarizes the above-described bar, the process of determining the pattern defect and the similar defect is similar to that caused by the protective film scratch when both the transmission illumination image inspection 151 and the side illumination image inspection 152a are shown as white defects. Since it is a defect, it is determined that light guide plate printing is a good quality product (153). The light guide plate is determined to be good even when it appears as a white defect in the transmission illumination image inspection 151 but is normal in the side illumination image inspection 152a (154). When it appears as a white defect in the transmission illumination image inspection 151 but appears as a black defect in the side illumination image inspection 152a, it is determined that it is a pattern defect, specifically, a defect in defection (155). If it is normal in the transmission illumination image inspection 151, it is immediately determined that it is good (154). If it appears as a black defect in the transmission illumination image inspection 151, but appears to be a white defect in the side illumination image inspection 152b, it is determined that it is a pattern defect, specifically an air defect. The light guide plate is determined to be good even when it is shown as a black defect in the transmission illumination image inspection 151 but normal in the side illumination image inspection 152b (154). If both of the transmission illumination image inspection 151 and the side illumination image inspection 152b appear as black defects, it is determined that the light guide plate is a good product because it is a similar defect caused by dust or bubbles (157).

When the pattern defect inspection method and apparatus according to the present invention is applied to the light guide plate pattern inspection, it is easy to control the quality of the product, and the defects of the surface foreign matter or the protective film which do not affect the quality of the product. Can be distinguished. This can ensure the reliability of inspection and improve the productivity of the product.

Although the detailed description of the present invention has been made, it will be apparent to those skilled in the art that various modifications may be made without departing from the scope of the present invention. The invention is only defined by the scope of the claims.

In the present invention, a pattern defect can be detected by classifying a real pattern defect and a similar defect provided by an inspection environment by acquiring a transmission illumination image and a side illumination image and comparing the defect candidates extracted from the two images. When the pattern defect inspection method and apparatus proposed in the present invention is applied to the light guide plate pattern inspection of the backlight unit, it is easy to control the quality of the product, and the surface defect or the protective film of the pattern defect does not affect the product quality. It can be distinguished from the above. This can ensure the reliability of inspection and improve the productivity of the product.

Claims (17)

In the pattern defect inspection method of the transparent plate having a periodic scattering pattern, (a) acquiring a transmission illumination image of the transparent plate using transmission illumination; (b) acquiring side lighting images of the transparent plate using side lighting; (c) extracting a first defect candidate that deviates from the periodicity of a pattern in the transmitted illumination image; (d) extracting a second defect candidate from the side illumination image that is inconsistent with the periodicity of the pattern; And and (e) detecting pattern defects by distinguishing between actual pattern defects and similar defects provided by an inspection environment by comparing the first defect candidates with the second defect candidates. The method of claim 1, wherein in steps (c) and (d), a spatial filter suitable for extracting a nonuniformity of a pattern through brightness differences of surrounding pixels spaced by a predetermined period is applied, and the spatial filter is convoluted with image data. And a portion corresponding to the image data is extracted as a defect candidate when the resulting filtering data exceeds a specific threshold by convolution. The inspection method according to claim 2, wherein the spatial filter is designed to emphasize high frequency components caused by pattern defects among image data converted to spatial frequencies based on periodicity of regular patterns. The method of claim 1 or 2, wherein the defect candidates brighter than the surrounding pixels spaced apart from the predetermined period are classified as white defects, and the candidates darker than the surrounding pixels spaced apart from the predetermined period are classified as black defects. The test method characterized by classifying as). The method of claim 4, wherein in step (e), Among the intersections of the first defect candidate and the second defect candidate, a defect candidate classified as a black defect in the transmission illumination image is classified as a white defect in the side illumination image, or a defect classified as a white defect in the transmission illumination image. And determining a defect candidate as a pattern defect only when a candidate is classified as a black defect in the side illumination image. The inspection method according to claim 5, wherein the step (d) is performed at the position of the first defect candidate in the side illumination image. According to claim 1, wherein the step (a), Irradiating light toward the transparent plate from the bottom of the transparent plate; And And acquiring the light transmitted through the transparent plate from an upper portion of the transparent plate. The method of claim 1 or 7, wherein the step (b), Irradiating light toward the transparent plate from both sides or one side of the transparent plate; And And acquiring scattered and reflected light in the transparent plate from above the transparent plate. The inspection method according to claim 8, wherein a reflection plate is disposed below the transparent plate to collect light scattered and reflected from the inside of the transparent plate to an upper portion of the transparent plate. In the pattern defect inspection apparatus of the transparent plate having a periodic scattering pattern, Transmissive illumination unit for irradiating light toward the transparent plate from the lower portion of the transparent plate; Side lighting units for irradiating light toward the transparent plate from both sides or one side of the transparent plate; Acquiring a side illumination image of the transparent plate by using the light irradiated from the transmissive illumination unit and transmitted through the transparent plate, and using the light reflected from the side illumination unit and scattered and reflected inside the transparent plate. An image acquisition unit; And Extracting a first defect candidate that deviates from the periodicity of the pattern in the transmitted illumination image, extracting a second defect candidate that deviates from the periodicity of the pattern in the side illumination image, and comparing the first defect candidate and the second defect candidate And an image processor for detecting pattern defects by distinguishing between actual pattern defects and similar defects provided by the inspection environment. 11. The method of claim 10, wherein the image processing unit comprises a spatial filter suitable for extracting the non-uniformity of the pattern through the difference in the brightness of the pixels around a certain period, and convoluting the spatial filter with the image data, the resulting filtering And when the data exceeds a specific threshold, a portion corresponding to the image data is extracted as a defect candidate. The inspection apparatus of claim 11, wherein the spatial filter is designed to emphasize high frequency components represented by pattern defects among image data converted to spatial frequencies based on a periodicity of a regular pattern. 12. The method of claim 10 or 11, wherein the image processing unit classifies the defect candidates brighter than the peripheral pixels spaced apart from the defect candidates as white defects and classifies the candidates darker than the peripheral pixels spaced apart from the predetermined cycles into black defects. Inspection device, characterized in that. The method of claim 13, wherein the image processor, Among the intersections of the first defect candidate and the second defect candidate, a defect candidate classified as a black defect in the transmission illumination image is classified as a white defect in the side illumination image, or a defect classified as a white defect in the transmission illumination image. Only when a candidate is classified as a black defect in the side illumination image, the defect candidate is determined as a pattern defect. The inspection apparatus of claim 10, wherein the image processor extracts the second defect candidate at a position of the first defect candidate. The inspection apparatus of claim 10, further comprising a reflector below the transparent plate to collect scattered and reflected light in the transparent plate above the transparent plate. The inspection apparatus according to claim 10, wherein the light sources of the transmissive illumination portion and the side illumination portion have a wavelength in the visible light region.

Images