JP2011209113A - Inspection system - Google Patents

Abstract

It is possible to reduce a confirmation work load by a microscope by processing defect images for all defects on a substrate with a plurality of thresholds to determine the type of defect and pass / fail, thereby improving work efficiency. Provide an inspection system that improves
An inspection system for detecting a defect in a substrate, wherein the substrate is transported based on defect detection information of a defect detected in advance by an automatic defect inspection apparatus during or after the manufacturing process of the substrate. A means for illuminating the substrate; a means for imaging the defect; and an image processing means for processing the image of the captured defect, and binarizing the processed defect image with a plurality of threshold values. A function for determining the same defect from the binarized position coordinates of the defect, and a function for determining the defect type of the defect determined to be the same defect and a function for determining pass / fail. Inspection system.
[Selection] Figure 12

Description

  The present invention relates to an inspection system for a color filter substrate that does not meet quality standards generated in a color filter manufacturing process used in a color liquid crystal display device.

  FIG. 1 is a cross-sectional view showing an example of a color filter used in a color liquid crystal display device. The color filter 1 includes a black matrix (hereinafter referred to as BM) 3, a red R colored pixel (hereinafter referred to as R pixel) 4-1, a green G colored pixel (hereinafter referred to as G pixel) 4-2, blue on a glass substrate 2. B color pixel (hereinafter referred to as B pixel) 4-3, transparent electrode 5, photo spacer (hereinafter referred to as PS) 6, and vertical alignment (hereinafter referred to as VA) 7 are sequentially formed. It is.

  As a method for manufacturing a color filter having the above structure, a photolithography method, a printing method, and an ink jet method are known. FIG. 2 is a flowchart showing steps of a commonly used photolithography method. In the color filter, first, a step of forming BM on the glass substrate (C-1), a step of cleaning the glass substrate (C-2), and a step of applying and pre-drying a colored photoresist (C-3) ), A pre-baking step (C-4) for drying and curing the colored photoresist, a step (C-5) for exposing, a step (C-6) for developing, and a step (C-) for curing the colored photoresist. 7) The process of forming a transparent electrode (C-8) and the process of forming PS and VA (C-9) are performed in this order.

  For example, when pixels are formed in the order of R pixel, G pixel, and B pixel, between the step of cleaning the color filter substrate (hereinafter referred to as the substrate) (C-2) and the step of curing the colored photoresist. In (C-7), the colored resist is changed in the order of red R, green G, and blue B, and the process is repeated three times to form R, G, and B pixels.

  The substrate to be manufactured is required to have high reliability. However, as described above, there are many steps in the manufacturing process of the color filter. In the manufacturing process of the color filter, FIG. 3 (about 5 μmφ to 500 μmφ on the color filter). The white defect 11 shown in a) and the foreign substance defect 12 shown in FIG. 3B occur. FIGS. 3A and 3B show examples of defects generated on the G pixel 4-2. In order to prevent the outflow of these defects, the defect is detected by an automatic defect inspection apparatus.

  FIG. 4 is a diagram showing an example of an automatic defect inspection apparatus used for detecting these defects. The automatic defect inspection apparatus includes a transport unit (air slider using rollers or air levitation transport) 23 for transporting the substrate 20 in the direction of an arrow 25, a reflection light source unit 21a and a transmission light source unit 21b for illuminating the substrate 20, and illumination. An image pickup unit 22 for picking up the imaged substrate 20 and a control unit 24 including an image comparison processing unit are provided. The reflection light source unit 21a, the transmission light source unit 21b, the imaging unit 22, and the transport unit 23 are connected to the control unit 24 and controlled.

The reflection light source unit 21a and the transmission light source unit 21b are installed so as to perform uniform illumination over the entire width in the direction orthogonal to the transport direction of the substrate 20, and light sources such as LEDs, halogen lamps, and xenon lamps are used. The imaging unit 22 includes a plurality of two-dimensional CCD cameras arranged with respect to the entire width in a direction orthogonal to the conveyance direction of the substrate 20. The two-dimensional CCD camera used here may be a monochrome camera, and the captured image shown in FIG. 5 is converted into, for example, 256 gradation digital data by an image data processing unit (not shown). The transport unit 23 may transport the substrate 20 with, for example, a transport roller 26, or a stage is provided for transport, the substrate 20 is placed on the stage, and the stage is moved and transported. Furthermore, the reflected light source unit 21a, the transmitted light source unit 21b, and the imaging unit 22 are moved while the substrate 20 is placed on a stationary mounting table to perform imaging.

  FIG. 5 shows an example of an image obtained by digitally converting a captured image and a defect. In general, a left / right / up / down repetitive pattern of one block of R pixel 4-1, G pixel 4-2, and B pixel 4-3 repeatedly provided at a cell pitch repeatedly provided at a cell pitch of a size of 100 μm to 6000 μm. Density comparison / difference processing is performed, binarization is performed with a set threshold value, defect portion 9 is extracted, and defect data (coordinates and pixel count) is stored in a file and database (hereinafter referred to as DB) by a computer including a personal computer.

  The size of the defect is replaced with the number of pixels of the CCD line sensor camera (hereinafter referred to as pixel number: unit PIX) that captured the defect, and this number of pixels is stored in the file and DB as a substitute value of the size of the defect. .

FIG. 6A shows the imaged defect 10a. FIG. 6B shows an example of a pixel 10b of a binarized defect CCD line sensor camera. In the case of FIG. 6B, the number of pixels of the CCD line sensor camera is 49 pixels in total in the X and Y directions, which is 49 pixels, and the defective pixel 10b is 11 PIX. For example the area of the defect pixel 10b because the pixels 10b in the area of the defect per 1PIX 10 * 10 = 100μm ² Tosureba the defect is a 11PIX becomes 11 * 100 = 1100μm ^2.

  The defect size, which is one of the above-described defect data, is used as a criterion for determining the quality of the defect. This defect size is obtained by counting the number of CCD pixels (pixel number: unit PIX) of the extracted defect portion. When the number of pixels exceeds a set numerical value (set value), the defective portion is treated as a defective defect (hereinafter, this set value is referred to as a determination size).

  For those treated as defective defects, the defect type is confirmed by a defect observing device or a visual judging device equipped with a microscope, and a final pass / fail judgment is performed.

JP 2009-41930 A

  Defects in the color filter manufacturing process include “resin-based foreign object defects” in which the colored photoresist material used to form the color pixels of the color filter is solidified, and “metal-based foreign object defects” due to dust generation from the manufacturing processing equipment. , "Human skin-related foreign object defect" generated from human body, "Fiber-based foreign object defect" generated from clothing, etc., "Droplet-based foreign object defect" generated from oil used in manufacturing processing equipment, dirt Various types of foreign matter defects and dirt defects such as “dirt defects” occur.

  Since these defects have different causes of occurrence, their appearances are also greatly different, but since the influence of quality on the actual product is also different, the judgment standard is different for each defect. For example, in the case of a resin foreign matter defect, it may be determined that a foreign matter defect smaller than a stain defect is defective, and conversely, even if the stain defect is larger than a resin foreign matter defect, it may be determined that there is no problem in quality. However, in the conventional method of performing density comparison difference processing as described above and binarizing with a set threshold value and extracting a defective portion, since these determinations are difficult, a defect observation apparatus having a microscope manually after completion of inspection The final confirmation was made using a visual judgment device and the quality was judged.

  When determining pass / fail by the defect observation apparatus or visual determination apparatus having the microscope, since the type of defect to be determined is not known, it is necessary to perform a pass / fail determination check operation for all defects, Therefore, it has been a cause of lowering work efficiency.

  Therefore, in view of such a problem, the present invention reduces the confirmation work load by a microscope by processing defect images for all defects in a substrate with a plurality of threshold values to determine the type of defect and pass / fail determination. It is an object to provide an inspection system that makes it possible to improve the work efficiency.

The invention according to claim 1 of the present invention is an inspection system for detecting a defect in a substrate.
Based on the defect detection information of the defects detected by the automatic defect inspection device in advance during or after the manufacturing process of the substrate,
Means for transporting the substrate;
Means for illuminating the substrate; means for imaging the defect;
Image processing means for processing the image of the imaged defect, and
A function of binarizing the processed defect image with a plurality of threshold values;
A function of discriminating the same defect from the binarized defect position coordinates;
An inspection system comprising: a defect type determination of a defect determined as the same defect and a function of determining pass / fail.

  The invention according to claim 2 of the present invention is such that the function of discriminating the same defect is the same defect when the distance of the position coordinates of the defect obtained by the plurality of threshold values is within a preset length. The inspection system according to claim 1, wherein the inspection system is discriminated.

  In the invention according to claim 3 of the present invention, the defect type discrimination and pass / fail judgment function includes a defect size binarized by a plurality of threshold values, an area ratio, and a plurality of arithmetic expressions. 3. The inspection system according to claim 1, wherein defect type discrimination and pass / fail judgment are made according to parameters.

The invention according to claim 4 of the present invention is an inspection system for detecting a defect in a substrate during or after the manufacturing process of the substrate,
Means for transporting the substrate;
Means for illuminating the substrate; means for imaging the defect;
Image processing means for processing the image of the imaged defect, and
A function of binarizing the processed defect image with a plurality of threshold values;
A function of discriminating the same defect from the binarized defect position coordinates;
An inspection system comprising: a defect type determination of a defect determined as the same defect and a function of determining pass / fail.

  According to the inspection system of the present invention, a defect detected by an automatic defect inspection apparatus is determined by a plurality of binarization thresholds, and a defect size is determined by defining a determination size for each defect type. Thus, it is possible to reduce the work load required for manually determining the type of defect and determining the quality of the defect, which has been conventionally performed. In addition, it is possible to perform stable inspection by eliminating variations in pass / fail judgment as compared with manual work.

The figure which showed an example of the color filter used for a color liquid crystal display device in a cross section. The flowchart of the process of the photolithographic method generally used. (A) is a figure which shows a white-out defect. FIG. 6B is a diagram showing the foreign object defect 12. The figure which shows an example of an automatic defect inspection apparatus. The figure which shows an example of the image and defect which digitally converted the imaged image. (A) is a figure which shows the imaged defect. (B) is a figure which shows an example of the pixel of the CCD line sensor camera of the binarized defect. The figure which shows schematic structure of the test | inspection system concerning this invention. (A) is a figure which shows the image containing the digitized defect concerning this invention. (B) is a figure which shows the image after a density comparison difference process. FIG. 6C is a diagram illustrating an image that has been binarized with respect to the image after the density comparison difference processing. The figure which shows an example of the method of adding the pixel part of the soiling defect which concerns on this invention. (A) is the schematic diagram of the image which digitized the picked-up dirt defect. FIG. 6B is a diagram illustrating an image binarized with a certain slice value. The figure for demonstrating the method to add. (A) is a figure which shows a dirt defect. (B) is a figure which shows the image which binarized (a). (C) is a figure which shows the image which added the image of (b). (D) is a figure which shows the stain | pollution | contamination defect in which addition was performed and the image process was carried out. The figure which shows three slice values based on this invention. The figure which shows three defect sizes at the time of setting the setting value of 3 levels based on this invention. (A) is a figure which shows the digitized image. (B) is a figure which shows the image after a density comparison difference process. FIG. 6C is a diagram showing an image of a defect binarized with three levels of slice values. The figure (a) which shows the schematic diagram for demonstrating the method to determine whether it is the same defect based on this invention is a figure which shows the defect obtained by slice value 1. (B) is a diagram showing a defect obtained by slice value 2; (C) is a figure which shows the defect obtained by the slice value 3. The figure for demonstrating the setting of the defect determination condition which concerns on this invention, and is a figure which shows an example of density distribution of a resin type foreign material defect and a dirt defect. (A) is a figure which shows density | concentration distribution of the resin type foreign material defect. (B) is a figure which shows the density distribution of a dirt defect. The figure which shows an example of the flow of determination of the kind of defect by the inspection system which concerns on this invention, and a quality determination.

  Hereinafter, an embodiment of an inspection system according to the present invention will be described with reference to the drawings.

  Substrate defects are detected by an automatic defect inspection apparatus in the manufacturing process and in the inspection process after the process is completed. The position information and size of the detected defect are stored in the file and DB as defect information.

  A schematic configuration of an inspection system according to the present invention is shown in FIG. The inspection system shown in FIG. 7 is illuminated by a transport unit 33 that is a transport unit for transporting the substrate 30 in the direction of the arrow 35, a reflection light source unit 31 a and a transmission light source unit 31 b that are illumination units that illuminate the substrate 30. It has an imaging unit 32 that is an imaging unit that images the substrate 30, an image processing unit 37 that is an image processing unit that processes the captured image, and a control unit 34. The reflected light source unit 31a, the transmitted light source unit 31b, the imaging unit 32, and the transport unit 33 are connected to the control unit 34 and controlled.

  The reflection light source unit 31a and the transmission light source unit 31b are installed so as to perform uniform illumination over the entire width in the direction orthogonal to the transport direction of the substrate 30, and light sources such as LEDs, halogen lamps, and xenon lamps are used. In the imaging unit 32, a plurality of two-dimensional CCD cameras are arranged with respect to the full width in the direction orthogonal to the conveyance direction of the substrate 30. The two-dimensional CCD camera used here may be for monochrome. The transport unit 33 may transport the substrate 30 with, for example, the transport roller 36, or may provide a stage for transport, place the substrate 30 on the stage, and move the stage. The reflection light source unit 31a, the transmission light source unit 31b, and the image pickup unit 32 may be driven while being mounted on a stationary table. Based on the defect position information obtained by the automatic defect inspection apparatus, the defect is imaged by the imaging unit 32 at that position.

  The image of the imaged defect of the substrate is digitized in, for example, 256 gradations by the image processing unit 37, and the R pixel 34-1 and the G pixel 34- are compared with the digitized image (FIG. 8A) including the defect 39. 2, the left / right / upper / lower repeating pattern of one group of B pixels 34-3 is subjected to density comparison difference processing (the image after the density comparison difference processing is shown in FIG. 8B), and then a set threshold value (0 to 255) is obtained. (Hereinafter referred to as slice value), binarization processing is performed on the image after the density comparison difference processing (the binarized image is shown in FIG. 8C). The slice value here means that the image after the density comparison difference processing is segmented on the basis of the slice value to perform video processing (in other words, the binarization level “0” or “1” on the basis of the slice value). Is processed).

  Pixel portions determined to be defective are extracted and added to calculate the defect size. A method of adding defective pixel portions will be described. FIG. 9 is a diagram illustrating an example of a method for adding a pixel portion having a stain defect. FIG. 9A is a schematic diagram of an image obtained by digitizing a captured dirt defect. An image obtained by binarizing this image with a certain slice value is shown in FIG. Generally, the stain defect has a higher density in the peripheral portion 40 than the density in the central portion 41. Therefore, the binarized image shown in FIG. 9B has a donut shape, and the binarization level of the peripheral portion is high. “1”, the binarization level of the central portion is “0”. The central portion of the image in FIG. 9B is set to “1” that is the same as the peripheral portion, that is, added to form an image of “1” in both the peripheral portion and the central portion as shown in FIG. Calculate the size.

  FIG. 10 is a diagram for explaining the schematic diagram of FIG. 9 in more detail. ((A), (b), and (c) in FIG. 9 and FIG. 10 are contrasted). The density value in the peripheral part of the stain defect shown in FIG. 10A is higher than that in the central part. When binarization is performed with the slice value shown in 42, only the peripheral part is binarized as shown in FIG. Becomes “1”, and the binarization level becomes “0” in the center. Further, addition is performed to obtain an image in which the binarization level of all parts from the peripheral part to the central part is “1” as shown in FIG. Thereafter, as shown in FIG. 10D, the stain defect is subjected to image processing.

  In the case where the slice value is 43 shown in FIG. 10A (in the case of the slice value 43 set to a level lower in density than the slice value 42), no addition is performed for this stain defect. This addition can be achieved by using a commonly used image processing technique.

  A function of binarizing the image-processed defect with a plurality of threshold values will be described. Assume that the slice value, which is a parameter required for the defect detection process, can hold slice values of N levels. For example, when the slice value of 3 levels is set, the set value is as follows: The above-described defect detection processing is performed for each slice value.

  That is, when setting values of three levels are set, the slice value 1 is S1, the slice value 2 is S2, and the slice value 3 is S3. In the case of the three slice values shown in FIG. 11, 44 is S1, 45 is S2, and 46 is S3. Reference numeral 51 denotes a defect density distribution.

  For example, when three levels of setting values are set for the defect 50 shown in FIG. 12, three defect sizes of the detected defect obtained by the processing are obtained. FIG. 12A shows a digitized image including the defect 50, and FIG. 12B shows a result after density comparison difference processing obtained by density comparison difference processing of the left, right, upper, and lower repeated patterns repeatedly provided at the cell pitch. Images are shown. FIG. 12C shows a binarized defect image with slice values set thereafter (slice values 42, 43, and 44 in FIG. 11). The defect size 1 obtained as a result of performing binarization processing on the slice value 1 is B1, and the defect size 2 obtained as a result of performing binarization processing on the slice value 2 is defined as defect size B2. The defect size 3 obtained by the binarized image obtained by binarizing the slice value 3 is defined as B3.

  In the above, the case where the slice values of three levels are set is illustrated, but the present invention is not limited to this, and the number of levels may be set as appropriate.

  A function for discriminating the same defect from the binarized defects will be described. All the defects in the detected substrate hold information indicating the plane position coordinates with respect to the substrate. For example, in the case of the defect 50 shown in FIG. 12, the center of gravity of the defect having the defect size of B1, B2, and B3. Information indicating the plane position coordinates of the position is assumed to be defect coordinates 1 (X1, Y1), defect coordinates 2 (X2, Y2), and defect coordinates 3 (X3, Y3), respectively.

  When the linear distances (B1 and B2, B2 and B3, B3 and B1) of the center of gravity of the defects of the defect sizes B1, B2, and B3 shown in FIG. 12 are within the set L (mm), respectively. Determines it as the same defect. The defects having the defect sizes B1, B2, and B3 shown in FIG. 12 are determined as the same defect because the positions of the centers of gravity are substantially the same. Even for the same defect, the positions of the centers of gravity are not always the same, and are often almost the same.

  Here, the reason for determining whether or not the defects of the defect sizes B1, B2, and B3 are the same defect is to give a defect ratio described later. The defect ratio is an area ratio in the same defect. As an example, FIG. 13 is a schematic diagram for explaining a method for determining whether or not the defects are the same. FIG. 13 (a) shows the defects 101 and 111 obtained by the slice value 1 and similarly FIG. 13 (b) shows the defects 102 and 112 obtained by the slice value 2 and FIG. Defects 103 and 113 obtained by the slice value 3 are shown. The center of gravity coordinates (X1, Y1) of the defect 101 are the distances from the center of gravity coordinates ((X2, Y2), (X20, Y20), (X3, Y3), (X30, Y30)) of the defects 102, 112, 103, 113. Is calculated. If the calculated result is within a preset L (mm), it is determined that the defect is the same. In the case of FIG. 13, as a result, the defect (101, 102, 103) is determined as the same defect. On the other hand, since the defect 101 is separated from the defects 112 and 113 by L (mm) or more, it is not determined to be the same defect. Similarly, the distance of the center of gravity coordinates of the defect 111 and the defects 102, 112, 103, 113 is calculated, and the defect of the result (111, 112, 113) is determined as the same defect.

The area ratio of the brute force is calculated for the defects determined as the same defect, and is stored and handled as a defect type determination parameter. For example, in the case of the example shown in FIG. 12, for the detected defect sizes of B1, B2, and B3, the area ratio is derived as a whole as shown below. That is,
As defect ratio 1 B12 = B1 / B2
B13 = B1 / B3 as defect ratio 2
As the defect ratio 3, the area ratio of B23 = B2 / B3 is calculated.

  The area ratio is calculated within the same defect.

  The function of defect type discrimination and pass / fail judgment will be described.

Parameters used for defect type determination and pass / fail determination are set in advance before inspection and recorded as determination condition recipes.
1) Mode name: Defect names to be determined, for example, “dirt defect”, “resin-based foreign object defect”.
2) Numerical threshold parameter: a determination threshold, which is a comparison reference value for defect ratio and defect size, for example, α, β, γ.
3) Logical operation parameters: parameters that are combined with the comparison method, for example, ≧, ≦,>, <, =, AND, OR.
4) Acceptance / rejection determination: Determines whether a defect conforming to the determination condition created with the above parameters is a non-defective product or a defective product, for example, OK or NG.

  A determination condition setting example will be described.

For example, if the mode name is resin-based foreign matter determination and the matching conditions are B12> α, And S1 ≧ β, AND S2 ≧ γ (that is, the defect ratio B12 is α or more, the defect size S1 is β or more, and The defect size S2 is equal to or larger than γ), and a defect conforming to the defect size S2 is determined to be NG as a resin-based foreign matter failure. Here, for example, α = 0.5, β = 500 μm ² , and γ = 1000 μm ² are set in advance.

As another example of setting the determination condition, if the mode name is determined to be dirty, and the matching conditions are B12 <α, And S1 ≧ β, AND S2 ≧ γ (that is, the defect ratio B12 is less than α, and the defect The size S1 is equal to or larger than β and the defect size S2 is equal to or larger than γ). Here, for example, α = 0.2, β = 1000 μm ² , and γ = 2000 μm ² are set in advance.

  As described above, by setting the numerical threshold parameter and the logical operation parameter in advance according to the feature of the defect, it is possible to determine the type of defect and determine the quality.

  The reason why the numerical threshold parameter and the logical operation parameter are set according to the defect feature will be described. FIG. 14 shows an example of the concentration distribution of the resin foreign matter defect and the soil defect shown above. FIG. 14A shows the concentration distribution of resin-based foreign matter defects, and FIG. 14B shows the concentration distribution of dirt defects. The concentration of the resin foreign matter defect has a high peak, and the concentration distribution gradient is large (B12 is large). On the other hand, the density of the dirt defect has a low peak and the density distribution gradient is small (B12 is small). In addition, since it is desirable that the resin-based foreign matter defect is a defect smaller than the stain defect (S1 and S2 is small), the defect judgment condition is set for each foreign matter defect or stain defect.

  In this way, the judgment condition settings are “resin-based foreign object defect”, “metal-based foreign object defect”, “human skin-based foreign object defect”, “fiber-based foreign object defect”, “droplet-based foreign object defect”, “dirt defect”, etc. It is set every time, it is determined whether or not all the defects meet the set conditions, defect type determination and pass / fail determination are performed, and the result is recorded as data.

  FIG. 15 shows an example of the flow of defect type determination and pass / fail determination.

After the manufactured color filter product substrate is put into the inspection system according to the present invention (J-1), the defect defect by the CCD camera is imaged based on the position information of the defect defect obtained by the automatic defect inspection apparatus. (J-2). Next, N defect detection processes are performed according to slice values of N levels, that is, slice values S1 to Sn (J-3). With respect to all the defects obtained by the slice values S1 to Sn, those having a linear distance of detected coordinates within L (mm) are determined as the same defect and recorded (J-4). The defect determined as the same defect is subjected to a matching process with a predetermined determination condition, and the defect type is determined and the quality is determined. The result is recorded as data (J-5), and then the product substrate is paid out (J-6).

  Although the inspection system according to the present invention is exemplified for the case of performing defect type determination and pass / fail determination based on defect position and defect size information detected in advance by an automatic defect inspection apparatus during or after the substrate manufacturing process. A system provided with the above function in an automatic defect inspection apparatus that inspects the substrate during the manufacturing process or after the manufacturing without passing through the automatic defect inspection apparatus. However, it is desirable to select which inspection system to use in consideration of the processing time of the inspection system.

  Moreover, although the case where the test | inspection system based on the said invention inspected a board | substrate was illustrated, it is not limited to this, It can apply to the test | inspection of a glass substrate widely.

  As described above, according to the inspection system of the present invention, it is possible to reduce the work load that has been required for manual defect type determination and defect quality determination. In addition, it is possible to perform stable inspection by eliminating variations in pass / fail judgment as compared with manual work.

DESCRIPTION OF SYMBOLS 1 ... Color filter 2 ... Glass substrate 3 ... Black matrix (BM)
4-1 ... Red R colored pixels (R pixels)
4-2 ... Green G coloring pixel (G pixel)
4-3 ... Blue B colored pixels (B pixels)
5 ... Transparent electrode 6 ... Photospacer (PS)
7 ... Vertical alignment (VA)
DESCRIPTION OF SYMBOLS 9 ... Defect part 11 ... White defect 12 ... Foreign object defect 20 ... Board | substrate 25 ... Arrow 23 which shows the direction which conveys a board | substrate 23 ... Conveyance part 21a ... Reflection light source part 21b ... Transmission light source unit 22 ... Image pickup unit 24 ... Control unit 26 ... Transfer roller 10a ... Image defect 10b ... Binary defect CCD line sensor pixel 30... Substrate 35... Arrow 33 indicating the direction in which the substrate is conveyed... Transport unit 31 a... Reflection light source unit 31 b. 34 ... Control unit 39 ... Digitized defect 34-1 ... R pixel 34-2 ... G pixel 34-3 ... B pixel 40 ... Peripheral part 41 of dirt defect ... The central part 42 of the stain defect ... the slice value 43 ... the slice value 4 Another slice value 44 ... slice value and its 1 (S1)
45 ... slice value 2 (S2)
46: Slice value 3 (S3)
51... Density distribution 50... Defects 101 and 111... Defects 102 and 112 obtained by slice value 1... Defects 103 and 113. Defect obtained by value 3

Claims (4)

An inspection system for detecting defects in a substrate,
Based on the defect detection information of the defects detected by the automatic defect inspection device in advance during or after the manufacturing process of the substrate,
Means for transporting the substrate;
Means for illuminating the substrate; means for imaging the defect;
Image processing means for processing the image of the imaged defect, and
A function of binarizing the processed defect image with a plurality of threshold values;
A function of discriminating the same defect from the binarized defect position coordinates;
An inspection system comprising: a defect type determination of a defect determined as the same defect and a function of determining pass / fail.   The function for discriminating the same defect is characterized by discriminating a case where the distance of the position coordinates of the defect obtained by the plurality of threshold values is within a preset length as the same defect. The inspection system described.   The defect type discrimination and pass / fail judgment function includes defect type discrimination and pass / fail judgment by a defect discrimination parameter including a defect size binarized by a plurality of threshold values, an area ratio, and a plurality of arithmetic expressions. The inspection system according to claim 1 or 2, characterized in that An inspection system for detecting defects in a substrate during or after the manufacturing process of the substrate,
Means for transporting the substrate;
Means for illuminating the substrate; means for imaging the defect;
Image processing means for processing the image of the imaged defect, and
A function of binarizing the processed defect image with a plurality of threshold values;
A function of discriminating the same defect from the binarized defect position coordinates;
An inspection system comprising: a defect type determination of a defect determined as the same defect and a function of determining pass / fail.

Images