CN110579477B

CN110579477B - Defect detection device and method of automatic repair system

Info

Publication number: CN110579477B
Application number: CN201810948050.0A
Authority: CN
Inventors: 安钟植; 尹汝凛; 丁炫硕; 朴世景; 金芝珉
Original assignee: Hb Technology Co ltd
Current assignee: Hb Technology Co ltd
Priority date: 2018-06-11
Filing date: 2018-08-20
Publication date: 2022-06-07
Anticipated expiration: 2038-08-20
Also published as: KR101982347B1; CN110579477A

Abstract

The invention discloses a defect detection device and method for a substrate. According to an aspect of the present invention, there is provided a defect detecting apparatus for detecting a defect of a substrate, comprising: a substrate transfer unit for transferring the substrate along a predetermined direction; an image generating section for generating an image of the substrate; a unit pixel recognition unit configured to recognize a unit pixel included in the substrate from an image of the substrate; a storage unit for storing a reference image of a unit pixel; an image comparing unit for comparing the generated image of each unit pixel with the reference image stored in the storage unit; a defect detection unit for detecting whether or not a defect exists in each unit pixel based on the comparison result; and a control unit for controlling each component in the defect detection device.

Description

Defect detection device and method of automatic repair system

Technical Field

The present invention relates to an apparatus and a method for detecting a defect existing in a substrate in an automatic Repair System (Auto Repair System).

Background

The statements in this section merely provide background information related to the present disclosure and may not constitute prior art.

With the remarkable development of display device technology, the technology related to image display devices of various types such as liquid crystal display and plasma display has been greatly advanced. In particular, in an image display device that realizes large-sized and high-precision display, a high degree of technological innovation is being advanced to reduce manufacturing costs and improve image quality. Glass substrates used for displaying images mounted on these various devices also require surface properties and conditions of higher dimensional quality and higher accuracy than ever before. In the production of glass for display devices and other applications, various production apparatuses are used to form glass substrates, but it is common practice to heat and dissolve inorganic glass raw materials, homogenize molten glass, and then form the glass into a predetermined shape. In this case, defects such as quality abnormality occur on the surface of the glass substrate due to various reasons such as insufficient melting of glass raw materials, unintended mixing of foreign materials during the manufacturing process, degradation of a molding apparatus, temporary molding conditions, and defects generated during cutting to a desired size after completion.

Various measures have been taken to suppress such defects generated in glass substrates, but it is difficult to completely prevent the generation of defects, and even if the generation of defects can be suppressed to some extent, if there is no technique for clearly identifying defective glass substrates, defective products that should be treated as defective products are mixed in glass substrates determined to be good products. Therefore, a technique for detecting defects of a glass substrate with high accuracy is very important.

As a method for detecting defects of a glass substrate, a visual inspection method depending on the sense of an inspector has been widely performed, but such a visual inspection method has limitations in terms of inspection accuracy and time required for inspection as the glass substrate is enlarged. Therefore, there is a need for developing an automatic inspection apparatus due to the limitation of a visual inspection method for inspecting defects of a glass substrate.

Disclosure of Invention

(problems to be solved by the invention)

It is an object of an embodiment of the present invention to provide an apparatus and method for analyzing an image of a substrate to detect defects present in the substrate.

(measures taken to solve the problems)

According to an aspect of the present invention, there is provided a defect detecting apparatus for detecting a defect of a substrate, comprising: a substrate transfer unit for transferring the substrate along a predetermined direction; an image generating section for generating an image of the substrate; a unit pixel recognition unit configured to recognize a unit pixel included in the substrate from an image of the substrate; a storage unit for storing a reference image of a unit pixel; an image comparing unit that compares the generated image of each unit pixel with the reference image stored in the storage unit; a defect detection unit for detecting whether or not a defect exists in each unit pixel based on the comparison result; and a control unit for controlling each component in the defect detection device.

According to an aspect of the present invention, the unit pixel identification unit identifies coordinates of each unit pixel.

According to an aspect of the present invention, the unit pixel identification unit identifies an area of each unit pixel using coordinates of each unit pixel.

According to an aspect of the present invention, the control unit sets a reference value, and the defect detecting unit determines that the detection target has a defect based on the comparison result using the reference value.

According to an aspect of the present invention, the control unit controls the unit pixel recognition unit such that the unit pixel recognition unit recognizes the unit pixels included in the substrate along a preset one direction and re-recognizes the unit pixels included in the substrate along another direction, thereby enabling the unit pixel recognition unit to recognize coordinates of the respective unit pixels.

According to an aspect of the present invention, there is provided a defect detection method, comprising: a process of generating an image of the substrate; a process of recognizing a unit pixel included in the substrate from an image of the substrate; a process of comparing the generated image of each unit pixel with a reference image of a pre-stored unit pixel; and a process of detecting whether or not a defect exists in each unit pixel based on the comparison result.

(Effect of the invention)

As described above, according to one aspect of the present invention, the present invention has an advantage in that defects existing within a substrate can be easily and accurately detected.

Drawings

Fig. 1 is a diagram illustrating a structure of a defect detecting apparatus according to an embodiment of the present invention.

Fig. 2 is a diagram illustrating a reference image of a unit pixel and an image of a substrate to be detected according to one embodiment of the present invention.

Fig. 3 is a diagram illustrating a result of analyzing a reference image and an image of a substrate to be inspected according to an embodiment of the present invention.

Fig. 4 is a diagram illustrating a result of analyzing a unit pixel image according to one embodiment of the present invention.

Fig. 5 is a diagram illustrating an area of each unit pixel within a substrate according to an embodiment of the present invention.

Fig. 6 is a diagram illustrating a reference image of a unit pixel, an image of a unit pixel, and a subtracted image according to one embodiment of the present invention.

FIG. 7 is a diagram illustrating a masking unit generated according to one embodiment of the present invention.

FIG. 8 is a diagram showing a method of generating a binarized image using a subtracted image and a masking unit according to one embodiment of the present invention.

Fig. 9 is a diagram showing the final binarized image after removing noise present in the binarized image according to an embodiment of the present invention.

Fig. 10 is a diagram illustrating a reference image of a unit pixel group that is variable according to an embodiment of the present invention.

Fig. 11 is a diagram illustrating a reference image of a unit pixel according to a second embodiment of the present invention.

Fig. 12 is a diagram illustrating a reference image of a unit pixel and an image of a substrate to be detected according to a second embodiment of the present invention.

Fig. 13 is a diagram illustrating an area of each unit pixel in a substrate according to a second embodiment of the present invention.

Fig. 14 is a flowchart illustrating a method of detecting a defect in a substrate by the defect detecting apparatus according to an embodiment of the present invention.

(description of reference numerals)

100: a defect detecting device; 110: a substrate transfer section; 120: an image generation unit;

130: a unit pixel identification unit; 140: a storage unit; 150: a pretreatment section;

160: an image contrast section; 170: a defect detection unit; 180: a control unit;

190: a communication unit; 210. 1010, 1020, 1030: a reference image of a unit pixel;

220: a substrate image; 225: an image of a unit pixel; 310. 320, and (3) respectively: line profile;

410: deducting the outline; 610: deducting image

Detailed Description

The present invention is capable of various modifications and embodiments, and specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. However, the present invention is not limited to the specific embodiments, and should be understood to include all modifications, equivalents, and alternatives included within the spirit and technical scope of the present invention. In the description of the drawings, like reference numerals are used for like constituent elements.

The terms first, second, A, B, etc. may be used to describe various components, but these components are not limited by these terms. The above terms are used only for the purpose of distinguishing one constituent element from another constituent element. For example, a first component may be designated as a second component, and similarly, a second component may be designated as a first component, without departing from the scope of the claims of the present invention. The term "and/or" includes a combination of a plurality of related items or one of a plurality of related items.

When a certain component is referred to as being "connected" or "connected" to another component, it is to be understood that another component may exist in the middle, although the component may be directly connected or connected to the other component. On the contrary, when a certain component is "directly connected" or "directly connected" to another component, it is to be understood that no other component exists therebetween.

The terminology used in the description presented herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The singular forms "a", "an" and "the" include plural forms as long as there is no definite other meaning in the context. In the present application, it is to be understood that the terms "includes" or "including" do not exclude the presence or addition of any feature, number, step, action, constituent element, component, or combination thereof described in the specification.

Unless otherwise defined, all terms including technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

Terms commonly used as defined in dictionaries should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

Fig. 1 is a diagram showing the configuration of a defect detecting apparatus according to an embodiment of the present invention.

Referring to fig. 1, a defect detecting apparatus 100 according to an embodiment of the present invention includes a substrate transfer section 110, an image generating section 120, a unit pixel identifying section 130, a storage section 140, a preprocessing section 150, an image comparing section 160, a defect detecting section 170, a control section 180, and a communication section 190.

The substrate for detecting defects includes a plurality of pixels (pixels). Typically, the substrate comprises a large number of pixels, each pixel also being composed of a plurality of layers (layers). The pixels in the substrate are realized by a plurality of layers such as an Active Layer (Active Layer), a Gate Layer (Gate Layer), a source Layer and a drain Layer. In the substrate creation process, the inflow of fine particles (particles) may cause defects in the substrate, and in particular, into any layer of pixels within the substrate, causing defects on a specific layer. The defect detection apparatus 100 can detect which pixel of which layer included in the substrate has a defect.

The substrate transfer part 110 is used to transfer a substrate in a predetermined direction. The substrate transfer part 110 transfers the substrate to an accurate position for detecting a defect. The substrate transfer unit 110 transfers the substrate having completed the defect inspection process, so as to perform a substrate processing process or a precision defect inspection process.

The image generating part 120 generates an image of the substrate in order to detect defects of the substrate. The image generating unit 120 is implemented by a device such as a camera capable of generating an image of the substrate, and generates an image of the substrate transferred by the substrate transfer unit 110.

The unit pixel recognition unit 130 recognizes the unit pixels included in the substrate from within the image of the substrate. The unit pixel recognition unit 130 analyzes the image of the substrate to recognize each unit pixel included in the substrate. The unit pixel recognition unit 130 analyzes the image of the substrate in the horizontal and vertical directions, and recognizes the start x-coordinate and the start y-coordinate of each unit pixel. The unit pixel identification unit 130 identifies the start x-coordinate and the start y-coordinate of each unit pixel, and can identify the x-axis length, the y-axis length, and the area of a specific unit pixel by using the start coordinate of the specific unit pixel and the start coordinate of a unit pixel adjacent to the specific unit pixel. For example, when the initial coordinate of a specific unit pixel is (0, 0) and the coordinates of the adjacent unit pixels are (45, 0) and (0, 45), the lengths of the specific unit pixel in the x-axis direction and the y-axis direction are 45 and 45, respectively, and the area of the specific unit pixel is 2025. In this manner, the unit pixel identification unit 130 identifies each unit pixel, and can determine whether or not a defect exists in a certain pixel when the presence of a defect is checked by comparing another configuration described later with the reference image.

The unit pixel identification unit 130 may identify each layer in the pixel at the same time when identifying each pixel in the substrate. Defects may occur in the entire pixel, but may also occur only in a portion of the layers within the pixel. In this case, if the unit pixel identification unit 130 identifies only the pixels, the defect detection device may not be able to accurately detect a specific position where a defect occurs in the pixels. Accordingly, the unit pixel identification part 130 may distinguish layers within the pixel by indicating an identifier or a color of each layer for the control part 180 to identify.

In addition, the unit pixel recognition part 130 may recognize the unit pixel using a Line Profile (Line Profile) method. The line profile method is a method of representing image characteristics by lines (lines). The line profile method may analyze only the entire image or a certain area or interval within the image, representing an average or sum associated with a certain direction. The unit pixel identifying part 130 may analyze such lines representing the image characteristics to identify the start coordinates of each unit pixel. However, the present invention is not limited to this, and any method may be used as long as the start coordinates of each unit pixel can be identified. For convenience of explanation, the unit pixel recognition unit 130 recognizes the unit pixel by using the line profile method.

The storage unit 140 stores a reference image of a unit pixel. The storage unit 140 stores the reference image of the unit pixel to determine whether or not a defect exists in each unit pixel by comparison with each unit pixel recognized by the unit pixel recognition unit 130. The reference image of the unit pixel stored in the storage unit 140 corresponds to an image of a unit pixel having no defect. The storage unit 140 stores an image of a unit pixel without any defect as a reference image, and detects whether or not a defect exists in each unit pixel by comparing the reference image with another structure described later.

The storage unit 140 can replace the stored reference image of the unit pixel under the control of the control unit 180. When the number of pixels included in the substrate is large, the pixels may have a difference according to the positions disposed in the substrate. For example, when the substrate is m × m for manufacturing a plurality of display devices, a relatively large number of pixels are included in the substrate. Therefore, even if there is no structural difference between the pixels located at one end of the substrate and the pixels located at the other end, slight differences in color, brightness, contrast, and the like occur. Based on this difference, there is a possibility that a unit pixel having no substantial (structural) defect is also detected as having a defect. In general, the probability that unit pixels having similar colors, brightness, contrast, and the like are arranged in a specific region in a substrate is high. Therefore, when the same reference image as one end of the substrate is selected, the unit pixels arranged at the other end of the substrate and the unit pixels arranged in the peripheral area thereof are different from the reference image, and it is possible to detect that there is a defect. In order to prevent such a problem, the control section 180 may replace the reference image of the unit pixel by a predetermined area. For example, the control section 180 replaces a preset unit pixel reference image with an image most similar to the reference image among images differing from the reference image. Therefore, the reference image of the unit pixel is replaced according to the area of the substrate, and the accuracy of detecting the defect may increase. The storage unit 140 may replace and store the reference image of the unit pixel or may additionally store the reference image according to the control of the control unit 180.

The preprocessing section 150 is used to preprocess the reference image of the unit pixel. The preprocessor 150 preprocesses the reference image of the unit pixel before the image comparison unit 160 compares the image of each unit pixel in the substrate with the reference image of the unit pixel so as to smoothly perform the comparison. For example, when the long axis of the image of each unit pixel is arranged along the x-axis direction and the long axis of the reference image of the unit pixel stored in the storage unit 140 is arranged along the y-axis direction, it is not convenient to compare the image of each unit pixel with the reference image of the unit pixel. To solve such inconvenience, the preprocessing section 150 may convert the direction of the reference image. The preprocessing section 150 may perform preprocessing such as direction conversion of the reference image so that the image comparison section 160 smoothly performs comparison of each unit pixel image and the unit pixel reference image.

The image comparing unit 160 compares the image of each unit pixel with the unit pixel reference image to check whether or not there is a defect in each unit pixel. When the image matching unit 160 matches the image of each unit pixel with the reference image of the unit pixel, the image matching unit may match only the image of the unit pixel having the area of the image of each unit pixel equal to or larger than the predetermined value set by the control unit 180. For example, the unit pixel images arranged at the respective ends in the image of the substrate may be generated in a state in which the image of the substrate generated by the image generating unit 120 is cut off, instead of being exposed to the entire area. In this way, since it is not difficult to compare an image of a unit pixel having an area of all the unit pixels with a reference image, the image comparing unit 160 may compare only an image of a unit pixel having an area of each unit pixel equal to or larger than a predetermined value.

The defect detection unit 170 detects a defect in each unit pixel using the result of the comparison by the image comparison unit 160. The defect detection unit 170 performs a luminance adjustment and binarization (binning) process on the image background of the contrast result of the image contrast unit 160, and finally detects whether or not a defect exists.

The control unit 180 controls the operations of the components (110 to 170 and 190). The control part 180 controls the unit pixel recognition part 130 to recognize the unit pixels within the image of the substrate along the image of the substrate in all directions including the horizontal direction and the vertical direction. The control unit 180 sets a numerical value for specifying a unit pixel to be compared so that the image comparison unit 160 compares the reference image with the image of each unit pixel. The control section 180 causes the image comparing section 160 to perform image comparison with the reference image only for the unit pixels having an area equal to or larger than a preset value. When excessive defects are detected based on the detection result of the defect detecting unit 170, the control unit 180 controls the storage unit 140 so that the reference image stored in the storage unit 140 is replaced with an image most similar to the reference image among unit pixel images different from the currently stored reference image, or an image most similar to the reference image is additionally stored. Thus, the control unit 180 can correct a fine difference between the unit pixels due to the arrangement of the unit pixels in the substrate. Further, the control part 180 may control the defect detection level of the defect detection part 170. For example, when the defect detecting apparatus 100 is used in an environment where no minute defects of the substrate are tolerated, the control part 180 may control the defect detecting part 170 so that even minute differences (for example, differences in color, contrast, etc., rather than structural differences) are detected as defects. In contrast, when the defect detecting apparatus 100 is used in an environment in which only defects on the structure within the unit pixel are detected, the control section 180 may control the defect detecting section 170 so as not to detect a minute difference as a defect. The above is shown in fig. 10.

Fig. 10 is a diagram illustrating a reference image of a unit pixel group varied according to one embodiment of the present invention.

The storage unit 140 may replace the reference image 210 pre-stored in the control unit 180 with one or more of the

other images

1010, 1020, and 1030, or additionally store one or more of the

other images

1010, 1020, and 1030. As shown in fig. 10, the

images

1010, 1020, 1030, which are substituted for the reference image 210 or additionally stored, are not structurally different from the reference image 210, and have only slight differences in contrast, color, and the like.

The control unit 180 controls the communication unit 190 so that the coordinates of the unit pixel detected as having the defect by the defect detection unit 170 are transmitted to an external device (not shown). The control unit 180 transmits the coordinates of the defective unit pixel to an external device (not shown) for repairing (repairing) the defect, thereby repairing the defect. Alternatively, the control unit 180 may transmit the coordinates of the defective unit pixel to an external device (not shown) for more accurately detecting the defect, thereby more accurately detecting the defect on the substrate. The in-substrate defect may exist in a specified area at or off the center of the substrate and thus may be detected in the image of the substrate. However, in addition to this, there may be a case where a defect in the substrate exists in the periphery of the substrate and cannot be detected in its entirety in the image of the substrate. In the latter case, since the defect existing in the substrate is not completely detected, when a process of repairing the defect is performed immediately, the same process of repairing the defect may be performed again for the remaining defect, thereby causing inconvenience. In order to prevent such inconvenience, the control part 180 transmits the coordinates of the unit pixel to an external device (not shown) for more precisely detecting the defect, so that the external device precisely detects again which region the defect exists up to with the coordinates as a center (the center portion of the substrate). Thereby, all areas of the defect present in the substrate are detected.

The communication unit 190 transmits the coordinates of the unit pixel in which the defect is detected to an external device.

Part (a) of fig. 2 illustrates a reference image 210 of a unit pixel stored in the storage section 140 according to an embodiment of the present invention. The reference image 210, which is an image of a unit pixel without any defect, may be stored in the storage part 140 before the defect detection apparatus 100 detects a defect of the substrate, or may be replaced in the process of the defect detection apparatus 100 detecting a defect of the substrate.

Part (b) of fig. 2 shows an image 220 of the substrate generated by the image generation section 120 according to one embodiment of the present invention. The image 220 of the substrate includes an image 225 of each unit pixel.

As shown in fig. 2, the reference image 210 and the image 225 of each unit pixel may be arranged in different directions. In this case, it is substantially difficult to detect whether or not a defect exists in each unit pixel within the substrate, and therefore, the preprocessing section 150 converts the direction of the reference image 210 so as to be arranged along the same direction as the image of each unit pixel. According to the example shown in fig. 2, the preprocessing section 150 may rotate the reference image 210 clockwise by 90 degrees such that the major axis of the reference image 210 is arranged along the y-axis direction.

Fig. 3 is a diagram illustrating a result of analyzing a reference image and an image of a substrate to be inspected according to one embodiment of the present invention.

The unit pixel recognition part 130 analyzes the reference image and the substrate image, respectively. The unit pixel recognition unit 130 may analyze only a predetermined section of the reference image and calculate an average value with respect to a predetermined direction. As a result, the unit pixel recognition part 130 can generate a line profile of the reference image. On the other hand, the unit pixel identifying section 130 generates a line profile 310 of an image for the substrate in the same manner. However, since a large number of unit pixels are included in the substrate, an average value of images of the substrate is calculated at every preset interval. Here, the preset interval may be set to the length of the unit pixel. As shown in fig. 3, when the unit pixel recognition part 130 analyzes the image of the substrate along the vertical direction in order to recognize the start y-coordinate of each pixel, the preset interval may be set to the lateral length of the unit pixel. When the unit pixel recognition part 130 calculates the average value of the images of the substrates for one column, the unit pixel recognition part 130 moves by a preset interval to calculate the average value of the images of the substrates. Accordingly, the unit pixel recognition part 130 generates a line profile 320 for each column of the image of the substrate.

The unit pixel recognition part 130 re-analyzes the reference image and the image of the substrate in a direction perpendicular to the analysis direction of the reference image and the image of the substrate. The unit pixel recognition unit 130 recognizes not only the initial y coordinate but also the initial x coordinate in order to recognize the coordinates of each unit pixel in the image of the substrate. Therefore, the unit pixel recognition part 130 re-analyzes the reference image and the image of the substrate in a direction perpendicular to the direction in which the analysis of the reference image and the image of the substrate has been performed. The unit pixel recognition unit 130 generates line profiles 310 and 320 of a reference image and an image of a substrate, respectively, which are analyzed in the vertical direction.

Fig. 4 is a diagram illustrating an analysis result of a unit pixel image according to an embodiment of the present invention.

The unit pixel recognition unit 130 performs subtraction operation on the generated line profile 310 of the reference image and the line profile 320 of the substrate image. Since the line profile 310 of the reference image and the line profile 320 of the image of the substrate have different lengths, the unit pixel recognition unit 130 repeatedly subtracts the line profile 310 of the reference image from the line profile 320 of the image of the substrate to calculate a subtracted line profile 410.

The subtractive line profile 410 has very low values of

origin

420, 423, 426, 429 relative to the perimeter. The starting point occurs at the unit pixel and the boundary between the unit pixels, and the subtractive line profile 410 has a very low value at the starting point of the boundary between the unit pixels.

Since the unit pixel recognition unit 130 calculates the subtracted line profile for both the line profile generated in one direction for the reference image and the substrate image and the line profile generated in a direction perpendicular thereto, it is possible to recognize the initial x-coordinate and the y-coordinate of all the unit pixels included in the substrate image. In addition, since the coordinates of all the unit pixels can be recognized, the area of each unit pixel can be obtained together.

When the unit pixel recognition part 130 recognizes all the unit pixels within the image of the substrate and the preprocessing part 150 completes the preprocessing required for the reference image stored in the storage part 140, the image comparison part 160 compares the reference image with the image of each unit pixel within the image of the substrate. At this time, the image contrast section 160 checks the area of each unit pixel in the image of the substrate before performing image contrast. The image comparing unit 160 performs image comparison with the reference image only for the unit pixels having the area of each unit pixel equal to or larger than a predetermined value. For example, the unit pixel 510 or the unit pixel 515 is located at each end, and not all area is included in the image of the substrate. Since such a unit pixel cannot be completely compared, the image comparison unit 160 does not perform image comparison on such a unit pixel.

Referring to part (a) of fig. 6, the image comparing section 160 compares the image 225 of each unit pixel remaining after the preprocessing process described with reference to fig. 5 with the reference image 210. Since the image without any defects in the unit pixel image is substantially the same as the reference image 210, no image can be derived from the subtracted image. In contrast, as shown in fig. 6, when the image contrast section 160 subtracts the image 225 of the defective unit pixel and the reference image 210, a subtracted image 610 of only the defective portion is generated.

The image comparator 160 may generate the subtracted image 610, and may calculate an average value of the subtracted image 610. Since the unit pixel image 225 and the reference image 210 include R, G, B channels (channels), if the unit pixel image 225 and the reference image 210 are subtracted from each other for each channel, the amount of calculation in the subsequent steps may increase. Therefore, in order to further increase the calculation speed, the image comparing unit 160 may calculate and average subtraction values related to the R, G, B channels to generate the average subtraction image 610.

Referring to part (b) of fig. 6, similarly, the image matching unit 160 also calculates an average value of the reference image 210 to generate an average reference image 620. As described above, since the reference image 210 also includes the R, G, B channel, there is a fear that the amount of computation may increase when the reference image 210 is used by another structure later. Therefore, the image matching unit 160 also calculates the average value of each channel for the reference image 210 to generate the average reference image 620.

Mask units (masks) 710, 720 are formed based on the pattern (pattern) of the reference image for definitely grasping the subtracted image 610, particularly, the portion where the average subtracted image 610 is different from the reference image. Masking

units

710, 720 for binarizing the subtracted image 610, particularly, the average subtracted image 610, respectively use a masking unit 710 that leaves only a portion other than the pattern formed in the reference image 210 and a masking unit 720 that leaves only a portion of the pattern formed in the reference image.

The defect detecting section 170 generates a binary image 810 of the subtracted image by using the subtracted image 610 and the masking unit 710. The defect detecting unit 170 applies the masking unit 710 to the subtracted image 610, and leaves only a portion (portion having a defect) other than the pattern in the subtracted image 610. The defect detecting section 170 applies a threshold value to the subtracted image to which the masking unit 710 is applied, and generates a binarized image 810 of the subtracted image. The defect detecting section 170 applies a preset critical value to binarize the subtracted image to which the masking unit 710 is applied.

The defect detecting section 170 removes noise 910 present in the binarized image 810. When the thickness of the defect detected within the binarized image 810 is smaller than a preset size or smaller than a preset area, the defect detecting unit 170 may classify the defect as noise which is not a defect. The noise 910 is generally caused by a slight positional error between internal patterns in the process of generating the subtracted image 610 by the image contrast section 160 or the process of generating the binarized image 810 by the defect detection section 170 applying the masking unit 710. Since this noise 910 is not the object to be detected, the defect detecting section 170 removes the noise 910 present in the binarized image 810. The defect detecting section 170 detects a defect in the noise-removed binarized image 810.

At this time, the defect detecting unit 170 may determine the type of the detected defect. The types of defects include Dark (Dark) type defects and Seed (Seed) type defects. The dark type defect is a defect generated by inflow of particles or the like during substrate production as large as the area of the inflow factor. The seed type defect is a defect that occurs in a constant area around a factor in addition to the area of the factor that flows in. Only if the two can be distinguished, the area of the defect can be judged later, and the process of repairing the defect can be carried out in different modes according to the defect. The defect detecting section 170 may determine whether it is a dark type defect or a seed type defect in consideration of the ratio of the total area of the defects detected within the binarized image 810 to the area of the inflow factor.

The unit pixel of the reference image 1110 may include a plurality of pixels. The substrate includes a plurality of pixels, and the number of pixels constituting each layer of the pixels may be different. For example, in the active layer or the gate layer, one pixel may constitute a unit pixel, but in the source and drain layers, two or more pixels may constitute one unit pixel. In this manner, the unit pixels in each layer may be different, and the storage unit 140 may store the reference image 1110 for each layer in order to detect defects existing in various layers in order to detect defects on all layers. Hereinafter, for convenience, a case where two pixels are included in the unit pixel in the reference image 1110 will be described, but the present invention is not limited thereto.

The reference image 1110 may be divided into left and

right pixels

1114, 1118, but the area of each

pixel

1114, 1118 need not be the same. The areas of the

pixels

1114 and 1118 may be the same as each other, but may be different from each other. The area ratios of the

pixels

1114 and 1118 in the reference image 1110 are stored in the storage unit 140. For example, the size of the pixel 1114 may be (0.49, 1), the size of the pixel 1118 may be (0.51, 1), and the ratios may be stored in the storage unit 140 together.

The unit pixel recognition unit 130 recognizes each unit pixel included in the substrate image 1210, and after recognition, the image comparison unit 160 compares the reference image 1110 stored in the storage unit 140 with the substrate image 1210. Of course, the reference image 1110 may be subjected to a preprocessing process by the preprocessing section 150 before the image comparison section 160 performs comparison. The image contrast section 160 contrasts the image 1220 of the unit pixel in the image 1210 of the substrate with the reference image 1110. When the

images

1110 and 1220 are compared, the image comparison unit 160 performs comparison corresponding to the position of each pixel included in the unit pixel. That is, the image comparing unit 160 compares the left pixel 1114 on the reference image 1110 with the left pixel 1224 on the unit pixel image 1220, and compares the right pixel 1118 on the reference image 1110 with the right pixel 1228 on the unit pixel image 1220.

In this case, the area ratio of each pixel on the unit pixel image 1220 may be different from the area ratio of each pixel of the reference image 1110. At this time, the image comparing unit 160 multiplies the area ratio of the reference image stored in the storage unit 140 to calculate the size of the unit pixel image. Accordingly, the image comparing part 160 can easily compare the reference image 1110 and the unit pixel image 1220 having different area ratios.

Then, the defect detecting unit 170 detects a defect existing in each unit pixel using the image comparison result.

When the image comparing unit 160 compares the

images

1110 and 1220, it is determined whether or not the area of each pixel in the image 1220 of the unit pixel is equal to or greater than a predetermined value set by the control unit 180. When the area of any one of the

pixels

1224 and 1228 included in the image 1220 of the unit pixel does not exceed the preset value, the image comparing section 160 does not perform the comparison with the reference image 1110. Since it is difficult to compare images of unit pixels not showing all areas, the image comparison unit 160 grasps the areas of all pixels included in each unit pixel, and determines whether or not the area is equal to or larger than a predetermined value to select an image of each pixel to be compared.

Fig. 14 is a flowchart illustrating a method of detecting a defect in a substrate by the defect detecting apparatus according to one embodiment of the present invention.

The image generating unit 120 generates an image of the substrate to be detected (S1410).

The unit pixel recognition unit 130 recognizes each unit pixel included in the substrate within the image of the substrate (S1420). The unit pixel recognition unit 130 generates a line profile of the reference image and the image of the substrate by using a line profile method. When generating the line profile, the unit pixel recognition unit 130 generates the line profile for each of the image of the substrate and the reference image in all directions (one direction and a direction perpendicular to the one direction).

The unit pixel recognition unit 130 recognizes information of each unit pixel included in the substrate using the reference image (S1430). The unit pixel recognition unit 130 subtracts the line contour of the reference image and the line contour of each unit pixel to generate a subtracted line contour. The unit pixel identification unit 130 identifies the boundary of each unit pixel by subtracting the line outline, and identifies the (initial) coordinate of each unit pixel. Further, since the coordinates of each unit pixel are recognized, the unit pixel recognition unit 130 can recognize the length in the x-axis direction and the y-axis direction and the area of each unit pixel.

The image comparing unit 160 compares the image of each unit pixel with the reference image of the unit pixel (S1440). When the unit pixel recognition part 130 recognizes all the unit pixels within the image of the substrate and the preprocessing part 150 completes all the preprocessing required for the reference image stored in the storage part 140, the image comparison part 160 compares the reference image with the image of each unit pixel within the image of the substrate. The image comparison section 160 performs image comparison to detect whether or not a defect exists in each unit pixel.

The defect detection unit 170 detects a defect in each unit pixel using the comparison result (S1450).

Fig. 14 shows the respective processes as being performed in sequence, but this is merely an example of the technical idea of the embodiment of the present invention. In other words, a person skilled in the art to which an embodiment of the present invention pertains can apply various modifications and variations to the present invention without departing from the essential characteristics of the embodiment of the present invention by changing the order of the processes shown in the drawings or by performing one or more of the processes in parallel. Thus, fig. 14 is not limited to a chronological order.

On the other hand, the process shown in fig. 14 may be embodied as computer readable codes on a computer readable recording medium. The computer-readable recording medium includes all kinds of recording devices for storing data that can be read by a computer system. That is, the computer-readable recording medium includes magnetic storage media (e.g., read-only memory, floppy disks, hard disks, etc.), optical reading media (e.g., compact disk read-only memory, digital video disks, etc.), carrier waves (e.g., transmission through the internet), and the like. Furthermore, the computer-readable recording medium can also be dispersed over a network-connected computer system, and the code can be stored and executed by the computer in a dispersed manner.

The above description is merely an exemplary illustration of the technical idea of the present embodiment, and a person skilled in the art to which the present embodiment belongs can make various modifications and variations within a range not exceeding the essential characteristics of the present embodiment. Therefore, the present embodiment is not intended to limit the technical idea of the present embodiment, but to illustrate the present embodiment, and the scope of the technical idea of the present embodiment is not limited thereto. The scope of the present embodiment is to be interpreted by the following claims, and all technical ideas within the equivalent scope thereof are to be interpreted within the scope of the present embodiment.

Claims

1. A defect detecting apparatus for detecting a defect of a substrate, comprising:

a substrate transfer unit for transferring the substrate along a predetermined direction;

an image generating section for generating an image of the substrate;

a unit pixel recognition unit configured to recognize a unit pixel included in the substrate from an image of the substrate;

a storage unit for storing a reference image of a unit pixel;

an image comparing unit for comparing the generated image of each unit pixel with the reference image stored in the storage unit;

a defect detection unit for detecting whether or not a defect exists in each unit pixel based on the comparison result;

a control unit for controlling each component in the defect detection device; and

a preprocessor capable of converting the direction of the major axis of the reference image,

the unit pixel recognition unit generates line outlines of the reference image of the unit pixel and the image of the substrate, respectively, generates a subtraction line outline by subtracting the generated line outlines, recognizes a point having a predetermined value or less in the generated subtraction line outline as a boundary between the unit pixels,

before the image comparison unit compares the image of the unit pixel with the reference image of the unit pixel, when the direction of the long axis of the reference image of the unit pixel and the direction of the long axis of the image of the unit pixel are different from each other, the preprocessing unit converts the direction of the reference image of the unit pixel so that the long axes of the images are in the same direction,

the control unit replaces the reference image of the unit pixel stored in the storage unit with any one of the images of the unit pixels generated by the image generation unit for a predetermined area in order to improve the accuracy of detecting the defect,

the control unit replaces the reference image of the unit pixel with an image most similar to the reference image among images different from the reference image.

2. The defect detecting apparatus according to claim 1, wherein the unit pixel identifying unit generates a line profile for the image of the substrate along one direction and a direction perpendicular to the one direction to identify the coordinates of each unit pixel.

3. The defect detecting apparatus according to claim 2, wherein the unit pixel identifying unit identifies an area of each unit pixel by using coordinates of each unit pixel.

4. The defect detecting apparatus according to claim 1, wherein the control section sets a reference value, and the defect detecting section determines that the object to be detected has a defect based on the comparison result using the reference value.