JP4027905B2 - Defect classification apparatus and defect classification method - Google Patents

Description

  The present invention relates to a defect classification device and a defect classification method.

  In a manufacturing process of a semiconductor wafer or the like, a multilayer wiring pattern is formed by repeating a process such as film formation, resist coating, exposure, development, etching, and resist peeling a plurality of times. Early detection of defects and optimization of the manufacturing process are indispensable items in order to improve the yield of products completed through a great number of processes.

  For this reason, the surface of the specimen is visually inspected before the etching process in which the actual pattern is formed, and defects are detected and defective products are identified. Traditional efforts have been made to optimize the process.

  However, the visual inspection has a problem that the burden on the inspector is large, and there is a difference between individuals and a uniform result cannot be obtained. Therefore, it has been attempted to realize such surface inspection using an optical device.

  Japanese Patent Laid-Open No. 2003-168114 by the present inventor discloses a method of classifying defects by macro inspection in which a surface of a subject is inspected using a relatively low resolution image having a pixel resolution of 100 to 150 μm.

  The principle of this defect classification method will be described below with reference to FIG. First, interference and diffraction images of the subject surface are acquired as an inspection image 700 ((A) in FIG. 22). When there are unevenness, scratches, shots (defects caused by the exposure machine), etc. on the surface of the object, these are captured as unevenness 701, shots 702, and scratches 703 in the inspection image 700. This is compared with a non-defective image 704 (FIG. 22B), which is a standard non-defective wafer interference and diffraction image. Alternatively, the comparison may be made in the unit of section having the same pattern in the interference and diffraction images of the subject surface. Thereby, a difference image 705 ((C) in FIG. 22) between the inspection image 700 and the non-defective image 704 is obtained. When this is binarized, a defect area composed of pixels having a large luminance difference from the non-defective image 704 is obtained as a defect extraction image 706 ((D) in FIG. 22).

Since the outer shape of the extracted defect area shape information, particularly the shot defect, which is a representative defect caused by the exposure machine, is similar to the exposure section shape, the correlation value with the template 707 indicating the exposure section shape (the luminance between corresponding pixels) The sum of the absolute difference values of the difference is calculated, and the defect type of the extraction region is classified based on this correlation value ((E) of FIG. 22).
JP 2003-168114 A

  However, in the above-described method, when shot defects are continuous as shown in FIG. 23A, only the same correlation value as the defect area inside the unevenness 701 can be obtained for the defect area inside the shot defect. Is difficult to clearly classify (see FIG. 23B). Conventionally, a defect classification apparatus and a defect classification method that showed an effective solution to such a problem have not been proposed.

  The present invention has been made paying attention to such a problem, and the object of the present invention is to provide a defect classification apparatus and a defect classification method capable of accurately classifying defects even for defects caused by successive exposure machines. Is to provide.

In order to achieve the above object, according to a first aspect of the present invention, there is provided a defect classification apparatus for classifying defects in a subject subjected to exposure for forming a wiring pattern, after the exposure is performed. Imaging means for picking up an image of the object surface, defect area extracting means for extracting a defect area from the image, defect area connecting means for connecting the extracted defect area as a connected defect area by morphological processing, and the continuous Yuiketsu Recessed exposure compartment depends contour ratio calculating means for calculating for each connection defect region position are linked to the ratio of the contour along the exposure compartment side is set in advance within the image of the contour of the region, based on the ratio And classifying means for classifying defects.

According to a second aspect of the present invention, in the defect classification apparatus according to the first aspect , the connection by the defect region connecting means is based on a morphological process using a structural element having a size larger than the width between adjacent chips.

According to a third aspect of the present invention, there is provided a defect classification apparatus for classifying defects in a subject that has been exposed to form a wiring pattern, wherein an image of the surface of the subject after the exposure is performed. imaging means for imaging a defect region extracting means for extracting a defective area from the image, a defect area aggregation means for aggregating the extracted defect regions compartment subdivided exposure section, the defect compartment aggregating the defective area among the defects section connecting means for connecting the adjacent defective sections as connecting defect compartment was connected the proportion of the contour along the communication Yuiketsu Recessed exposure compartment side position in the image of the contour of the partition is set in advance Exposure section dependent contour ratio calculating means for calculating each connected defect section ; and classifying means for classifying defects based on the ratio.

The fourth aspect of the present invention, Te defect classification apparatus odor classifying defects in a subject exposed is performed to form a wiring pattern, before SL after the exposure has been performed of the subject surface An image capturing means for capturing an image, a defect section detecting means for inspecting the image in units of subdivisions of the exposure section, and detecting a section where a defect exists as a defect section, and an adjacent defect section among the defect sections are connected Defect section connecting means to be connected as a defect section, and exposure for calculating each connected defect section in which the ratio of the contours along the exposure section side whose position in the image is set in advance among the contours of the connected defect sections is connected Compartment-dependent contour ratio calculation means, and classification means for classifying defects based on the ratio.

According to a fifth aspect of the present invention, in the defect classification apparatus according to the third or fourth aspect , a section obtained by subdividing the exposure section is matched with a chip section.

According to a sixth aspect of the present invention, in the defect classification method for classifying defects in a subject that has been exposed to form a wiring pattern, an image of the surface of the subject after the exposure is performed. captured, the image from the extracted defect regions, wherein the extracted defect regions linked as a connecting defect area by morphology processing, exposure position in the image of the contour of the communicating Yuiketsu Recessed region is set in advance The ratio of the contours along the partition sides is calculated for each connected defect area, and the defects are classified based on the ratio.

According to a seventh aspect of the present invention, in the defect classification device according to the sixth aspect , the extracted defect regions are connected by a morphological process using a structural element having a size larger than the width between adjacent chips.

Further, an eighth aspect of the present invention, Te defect classification method odor classifying defects in a subject exposed is performed to form a wiring pattern, before SL after the exposure has been performed of the subject surface An image is picked up, a defect area is extracted from the image, the extracted defect area is aggregated into sections obtained by subdividing exposure sections, and adjacent defect sections among the defect sections obtained by consolidating the defect areas are connected as connected defect sections. Then, a calculation is performed for each connected defect section in which the proportions of the contours along the exposure section side where the position in the image is set in advance among the contours of the connected defect sections are connected, and the defects are classified based on the ratio. .

According to a ninth aspect of the present invention, in the defect classification method for classifying defects in a subject that has been exposed to form a wiring pattern, an image of the subject surface after the exposure is performed. captured, the image is examined in section units obtained by dividing the exposure compartment to detect a compartment there is a defect as a defect section, connected as a coupling defect compartment defects adjacent sections of the defect compartment Recessed the communication Yuiketsu Of the contours of the sections, calculation is performed for each connected defect section obtained by connecting the proportions of the contours along the exposure section side where the position in the image is set in advance, and the defects are classified based on the ratio.

According to a tenth aspect of the present invention, in the defect classification method according to the eighth or ninth aspect , a section obtained by subdividing the exposure section is matched with a chip section.

  According to the present invention, it is possible to classify the defects caused by the exposure machine with higher accuracy.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

(First embodiment)
FIG. 1 is a diagram showing the configuration of the defect classification apparatus according to the first embodiment of the present invention. The defect classification apparatus according to the first embodiment forms an image of the illuminator 101 that illuminates the subject 111, the band-pass filter 102 that limits the wavelength of illumination light from the illuminator 101, and the reflected light from the subject 111. Lens 103, CCD camera 104 as an imaging means for converting the formed subject image into an electrical signal, an image input board 105 for capturing an output signal from the CCD camera 104 as an image, and retention of image data , And a memory 106 used for processing of each means described later, a defect area extracting means 107 for extracting a defect area from the acquired image data, a defect area connecting means 108 for connecting the extracted defect areas, and a connected connection An exposure section-dependent contour ratio calculating unit 109 that calculates the ratio of the contour along the exposure section side of the contour of the area, and a defect based on the calculated ratio. It consists classification means 110. similar.

  Of the above configuration, the memory 106, the defect area extraction means 107, the defect area connection means 108, the exposure section dependent contour ratio calculation means 109, and the classification means 110 are realized by the PC 120. Here, the memory 106 is embodied as a memory in the PC 120, and the defect area extraction means 107, the defect area connection means 108, the exposure section dependent contour ratio calculation means 109, and the classification means 110 are embodied as a CPU in the PC 120. It becomes.

  Details of the defect classification method according to the present embodiment will be described below with reference to the drawings. In this defect classification apparatus, the wavelength of light from the illuminator 101 is limited by the band-pass filter 102 to irradiate the subject, and diffracted light (or interference light) reflected from the subject surface is imaged by the lens 103 to form a CCD. It is converted into an electric signal by the camera 104. This electrical signal is digitized through the image input board 105 and is taken into the calculation memory 106, which becomes an inspection image of the subject 111. FIG. 2 is a diagram schematically illustrating an example of such an inspection image 130. The inspection image 130 generally includes continuous shot defects 131, uneven defects 132, and the like.

  Next, the defect area extraction unit 107 extracts a defect area from the acquired inspection image 130. The following two methods can be considered as extraction methods. In the first method, in the inspection image 130 as shown in FIG. 3A, a threshold value that is within a non-defective level luminance range is set, and a pixel region having a luminance exceeding this threshold value is determined as a defect. Extracted as an extracted image 131 (FIG. 3B). Here, the threshold value indicating the brightness range of the non-defective product level may be preset in the PC 120 or may be determined adaptively based on the brightness histogram in the image (The University of Tokyo Press: Image Analysis). (See Handbook: Mikio Takagi, Yoshihisa Shimoda, Supervision: 502P, Binarization).

  In the second method, a non-defective wafer image 704 as shown in FIG. 22B (or an image 140 of a certain section that becomes a non-defective product as shown in the upper part of FIG. 4A) is held, and this image and An inspection image 130 (or a corresponding section in the inspection image) as shown in FIG. 4A is aligned, a luminance difference between overlapping pixels is obtained, and a difference image 132 (FIG. 4B) is created. Using the difference image 132, a defective area is extracted by threshold processing similar to the first method.

  After extracting the defective areas in this way, the defective areas are connected by the defective area connecting means 108. For the connection at this time, closing which is one of the morphological processes is used (for details of the morphological processes, see Corona: Morphology: Hidefumi Obata).

Hereinafter, the morphology processing will be briefly described. The basis of the morphological process is two processes, dilation and erosion, which are performed on the target area A using a reference graphic (reference image) called a structural element B, respectively. Each process is expressed as follows.

  Hereinafter, each of the dilation process and the erosion process will be described with reference to the drawings. FIG. 5 is a diagram for explaining dilation processing. An area that can be covered by the structural element B when the origin B1 of the structural element B is translated in the target area A is output as a processing result C.

  FIG. 6 is a diagram for explaining the erosion process. An area that can be covered by the origin B1 of the structural element B when the structural element B is translated in the target area A is output as a processing result C.

There is a closing process as a process that is a combination of the above-described dilation process and erosion process. This process is expressed as follows.

  FIG. 7 is a diagram for explaining the procedure of the closing process. In the closing process shown in FIG. 7, when the structural element B is moved from the outside of the target area A, the area surrounded by the structural element B is output.

  FIG. 8 shows a result of closing processing of the defect extraction area 150 (FIG. 8A) described in FIG. 3 using a rectangular structural element 151 (FIG. 8B). FIG. 9 shows a result (FIG. 9B) obtained by closing the defect extraction region 150 (FIG. 9A) described with reference to FIG.

  FIG. 10 is a diagram for explaining the relationship between the width between adjacent chips and the structural element size. The area corresponding to the area between adjacent chips in the wafer (near the dicing line) has a different pattern state from the normal in-chip pattern. Therefore, the brightness in the inspection image is different from that in the chip. It becomes a main area that divides the defective area. Therefore, as shown in FIG. 10, by setting the size 161 of the structural element to be larger than the width 160 between the adjacent chips, the defect regions can be effectively connected.

  After the defective areas are connected in this way, the exposure section-dependent contour ratio calculating unit 109 calculates the ratio of the contours along the exposure section sides among the connected contours of the defective areas. This calculation procedure will be described below.

1) First, the outline of the connected defect area is extracted. There are various methods for contour extraction. Here, a method in the case of using a binary image in which the luminance value of the defective area = 1 and the luminance value of other pixels = 0 is shown. In this case, the product of the pixels 171-1 to 171-4 located in the vicinity of 4 around the target pixel 170 as shown in FIG. 11 is obtained at an arbitrary target pixel position in the defect area. By extracting the target pixel at the position, a contour portion is obtained. When the outline of the connection defect region shown in FIG. 12A (corresponding to FIG. 9B) is extracted by the above method, the result shown in FIG. 12B is obtained.

2) Next, a contour along the exposure partition side is obtained. The position of the exposure section in the image is set in advance. If the inspection image is shifted every time it is imaged, the exposure section position in the representative image serving as a position reference is set, and after aligning each inspection image with the representative image, the set exposure section position is set. get. Next, an image showing the range of the width X around the exposure partition side is created ((A) of FIG. 13). By taking the AND of this image and the contour image shown in FIG. 13B (corresponding to FIG. 12B), the contour along the exposure partition side as shown in FIG. 13C can be obtained. it can. If the width X is determined based on the section size such as X = α × “exposure section size”, it can be determined adaptively to the section size.

3) Of the contours of each connected defect area, the ratio of the contours along the exposure partition side is obtained.

  The ratio value obtained as described above becomes higher for a defect whose defect outline is similar to that of the exposure section, that is, a defect caused by the exposure machine. After the ratio value of the contour along the exposure section side is calculated, the defect area is classified based on the ratio value obtained by the classification unit 110. Here, defects caused by the exposure machine are classified. This is realized by setting a predetermined threshold value for the calculated ratio value, and if the ratio value is larger than this threshold value, the defect is attributed to the exposure apparatus, and if the ratio value is smaller, it is classified into others.

  FIG. 14 is a flowchart showing the procedure of the defect classification method described above. First, the subject is imaged by the CCD camera 104 to obtain an inspection image (step S1). Next, a defect area is extracted from the acquired inspection image by the defect area extraction means 107 (step S2). Next, the extracted defect areas are connected by the defect area connecting means 108 (step S3). Next, the contour of the connected defect area is extracted by the exposure section dependent contour ratio calculating means 109 (step S4), and the contour along the exposure section side is extracted from the extracted defect area contour (step S5). . Next, the exposure section dependent contour ratio calculating means 109 obtains the ratio of the contour along the exposure section side out of the contour lengths of the connected defect areas, that is, the contour length along the exposure section side / the total contour length (step S6). ). Next, the defect area is classified by the classification means 110 based on the obtained ratio value (step S7). Finally, the classification result is output (step S8).

  According to the first embodiment described above, the defect areas are connected, and the defects are classified based on the relationship between the outlines of the areas and the exposure partition sides, so that even with respect to the defects caused by successive exposure machines, the accuracy can be improved. Classification can be performed.

(Second Embodiment)
The second embodiment of the present invention will be described below. The defect classification apparatus according to the second embodiment of the present invention consolidates and connects defect areas extracted by the same method as the defect classification apparatus of the first embodiment into sections obtained by subdividing the exposure sections. The defect is classified based on the relationship with the exposure section.

  FIG. 15 is a diagram showing the configuration of the defect classification apparatus according to the second embodiment of the present invention. The configurations of 201 to 207 and 211 shown in FIG. 15 are the same as those of 101 to 107 and 111 of FIG. 1 showing the defect classification apparatus of the first embodiment, so description thereof will be omitted here. The defect classification apparatus according to the second embodiment further includes a defect area aggregating unit 208 for aggregating the extracted defect areas into areas obtained by subdividing the exposure areas, a defect area linking unit 209 for linking defect areas obtained by aggregating defect areas, An exposure section-dependent contour ratio calculating unit 210 that calculates the ratio of the contours along the exposure section side in the contours of the connected defect sections, and a classification unit 212 that classifies the defects based on the ratio.

  In the defect classification apparatus according to the first embodiment, it is a time consuming process to connect extracted defect regions in units of pixels or to extract contours. Therefore, in the defect classification apparatus according to the second embodiment, the extracted defect areas are aggregated into areas obtained by subdividing the exposure areas in the defect area aggregating unit 208, thereby reducing the subsequent calculation amount.

  As an aggregation method here, as shown in FIG. 16, a subdivision section obtained by subdividing the exposure section 250 into 1 / N (here, 1/16) is considered. Next, as shown in FIG. 17, subdivisions in which the defect area 260 exists are aggregated by making them into defect divisions. FIG. 17B is a diagram showing the result of collecting the extracted defect areas into subdivisions.

  After the defect areas are aggregated and the defect sections are obtained, the defect sections are connected by the defect section connecting means 209. This is dealt with by regarding a defective section adjacent in the vertical and horizontal directions as one connected defective section. After the defective sections are connected, the exposure section-dependent contour ratio calculation unit 210 calculates the ratio of the contours along the exposure section sides of the connected defective section contours. As a procedure, the same processing as that performed in the first embodiment is performed by regarding the subdivision as one pixel.

  18A shows a contour extraction result obtained from the aggregation result shown in FIG. 17B, and FIG. 18B shows a contour extraction result along the exposure section.

  The obtained ratio value is higher for a defect having a defect profile similar to that of the exposure section, that is, a defect caused by the exposure machine. After the ratio value is calculated, the classifier 212 classifies the defects caused by the exposure machine. This is realized by setting a predetermined threshold value for the calculated ratio value, and if the ratio value is larger than this threshold value, the defect is attributed to the exposure apparatus, and if the ratio value is smaller, it is classified into others. At the time of connecting defect sections, the area of the connected defect sections is obtained. If this value is larger than the exposure section size, the classification according to this method is used, and if the value is smaller, the classification based on the correlation value shown in JP-A-2003-168114 is used. By using it, it is possible to classify defects caused by the exposure machine with higher accuracy.

  FIG. 19 is a flowchart showing the procedure of the defect classification method described above. First, the subject is imaged by the CCD camera 204 to obtain an inspection image (step S11). Next, a defect area is extracted from the inspection image acquired by the defect area extraction unit 207 (step S12).

  Next, the defect area aggregating means 208 aggregates the defect area into sections obtained by subdividing the exposure sections (step S13). Next, in the defect section connecting means 209, sections (defect sections) in which defects are aggregated are connected (step S14). Next, the exposure section dependent contour ratio calculating means 210 extracts the contours of the connected defect sections (step S15), and obtains the contour along the exposure section side from the extracted contours (step S16). Further, the exposure section dependent contour ratio calculating means 210 calculates “contour length along the exposure section side / total contour length” (step S17). Next, the classifying unit 212 classifies the defect area based on the obtained value (step S18). Finally, the classification result is output (step S19).

  According to the second embodiment of the present invention, the defect area is aggregated and connected to sections obtained by subdividing the exposure section, and the defects are classified based on the relationship between the outline of the area and the exposure section side, thereby processing speed. While improving the above, it is possible to classify defects caused by successive exposure machines.

(Third embodiment)
The defect classification apparatus according to the third embodiment of the present invention inspects an inspection image obtained by the same method as the defect classification apparatus according to the first and second embodiments, and inspects the defect section by subdividing the exposure section. These defect sections are connected, and the defects are classified based on the relationship between the contour and the exposure section.

  FIG. 20 is a diagram showing the configuration of the defect classification apparatus according to the third embodiment of the present invention. The configurations of 301 to 306 and 311 shown in FIG. 20 are the same as 101 to 106 and 111 of FIG. 1 showing the defect classification apparatus according to the first embodiment, and a description thereof will be omitted here.

  The defect classification apparatus according to the third embodiment has a configuration that simplifies the process of extracting the defect area and consolidating the exposure sections into subdivided sections in the defect classification apparatus according to the second embodiment. That is, in the third embodiment, the defect section detection unit 307 performs inspection in units of subdivisions of the exposure section, detects a section having an abnormality as a defect section, and connects the detected defect section by the defect section connection unit 308. Then, the ratio of the contours along the exposure section side among the contours of the connected defect sections is calculated by the exposure section-dependent contour ratio calculation means 309, and the defect is classified by the classification means 310 based on the calculated ratio.

  Here, the defect section connecting means 308, the exposure section dependent contour ratio calculating means 309, and the classifying means 310 are the defect section connecting means 209, the exposure section dependent contour ratio calculating means 210, and the classifying means in the second embodiment (FIG. 15), respectively. This is equivalent to 212.

  As a method for detecting a defective section, image statistics (average brightness, brightness standard deviation, brightness histogram, brightness gradient histogram, etc.) of subdivided sections that are non-defective products are stored in advance. There is a method in which image statistics of subdivisions to be inspected corresponding to the above are compared, and a section whose difference exceeds a specified value is defined as a defective section. At this time, if there is a pattern commonality between the subdivision sections in the exposure section, it is possible to share the statistics. For this reason, when there are about 4 × 4 or more chip sections in one shot exposure section, the processing can be further simplified by matching the sub-section sections with the chip sections.

  FIG. 21 is a flowchart showing the procedure of the defect classification method described above. First, the subject is imaged by the CCD camera 304 to obtain an inspection image (step S21). Next, the defect section detection means 307 detects the defect section (step S22). The following steps S23 to S28 are the same as steps S14 to S19 of FIG. 19 which is the second embodiment.

  As described above, according to the third embodiment of the present invention, the subdivisions that are directly defective are detected and connected based on the inspection image, and the defects are classified based on the relationship between the contour of the region and the exposure side. Thus, it is possible to classify defects caused by successive exposure machines while improving the processing speed.

It is a figure which shows the structure of the defect classification apparatus which concerns on 1st Embodiment of this invention. It is a figure which shows an example of the test | inspection image 130 typically. It is a figure for demonstrating the defect area extraction by binarization. It is a figure for demonstrating the difference image creation by the comparison of the same pattern division in a test | inspection image. It is a figure for demonstrating dilation processing. It is a figure for demonstrating erosion processing. It is a figure for demonstrating a closing process. It is a figure explaining the procedure which connects the defect extraction area | region demonstrated in FIG. 3 using a rectangular structural element. It is a figure explaining the procedure which connects the defect extraction area | region demonstrated in FIG. 3 using a circular structural element. It is a figure which shows the relationship between the width | variety between adjacent chips, and structural element size. It is a figure for demonstrating the 4 neighboring pixel of an attention pixel. It is a figure for demonstrating the procedure which extracts the outline of a connection defect area | region. It is a figure for demonstrating the extraction procedure of the outline along an exposure division edge. It is a figure for demonstrating the flow of the defect classification | category process which concerns on 1st Embodiment of this invention. It is a figure which shows the structure of the defect classification device which concerns on 2nd Embodiment of this invention. It is a figure which shows the division which subdivided the exposure division of the test | inspection image shown in FIG. 2 into 1/16. It is a figure which shows the result of having condensed the extracted defect area | region into the subdivision division. It is a figure which shows the processing result based on the aggregation result of FIG. It is a figure for demonstrating the flow of the defect classification | category process which concerns on 2nd Embodiment of this invention. It is a figure which shows the structure of the defect classification device which concerns on 3rd Embodiment of this invention. It is a figure for demonstrating the flow of the defect classification | category process which concerns on 3rd Embodiment of this invention. It is a figure for demonstrating the principle of the conventional defect classification device. It is a figure which shows the example of application of the conventional method with respect to the shot defect which generate | occur | produced continuously.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 ... Illuminator, 102 ... Band pass filter, 103 ... Lens, 104 ... CCD camera, 105 ... Image input board, 106 ... Memory, 107 ... Defect area extraction means, 108 ... Defect area connection means, 109 ... Exposure zone dependent contour Ratio calculation means, 110 ... classification means, 120 ... PC.

Claims (10)

In a defect classification apparatus for classifying defects in a subject that has been exposed to form a wiring pattern,
Imaging means for capturing an image of the subject surface after the exposure is performed;
A defect area extracting means for extracting a defect area from the image;
Defect region connecting means for connecting the extracted defect regions as connected defect regions by morphological processing;
An exposure section-dependent contour ratio calculating means for calculating each connected defect area connecting the ratios of the contours along the exposure section side where the position in the image is set in advance among the contours of the connected defect areas;
A classifying means for classifying defects based on the ratio;
A defect classification apparatus comprising: The defect classification apparatus according to claim 1, wherein the connection by the defect area connecting means is based on a morphological process using a structural element having a size larger than the width between adjacent chips. In a defect classification apparatus for classifying defects in a subject that has been exposed to form a wiring pattern,
Imaging means for capturing an image of the subject surface after the exposure is performed;
A defect area extracting means for extracting a defect area from the image;
A defect area aggregating means for aggregating the extracted defect area into sections obtained by subdividing an exposure section;
A defect section connecting means for connecting adjacent defect sections as a connected defect section among the defect sections in which the defect areas are aggregated; and
An exposure section-dependent contour ratio calculating means for calculating, for each connected defect section, the ratio of the contours along the exposure section side where the position in the image is preset among the contours of the connected defect sections;
A defect classifying apparatus comprising: classifying means for classifying defects based on the ratio. In a defect classification apparatus for classifying defects in a subject that has been exposed to form a wiring pattern,
Imaging means for capturing an image of the subject surface after the exposure is performed;
A defect section detection means for inspecting the image in units of subdivisions of the exposure section, and detecting a section where a defect exists as a defect section;
A defect section connecting means for connecting adjacent defect sections among the defect sections as a connected defect section;
An exposure section-dependent contour ratio calculating means for calculating, for each connected defect section, the ratio of the contours along the exposure section side where the position in the image is preset among the contours of the connected defect sections;
A classifying means for classifying defects based on the ratio;
A defect classification apparatus comprising: 5. The defect classification apparatus according to claim 3, wherein a section obtained by subdividing the exposure section is matched with a chip section. In a defect classification method for classifying defects in a subject that has been exposed to form a wiring pattern,
Taking an image of the subject surface after the exposure has been performed,
Extracting a defect area from the image,
Connecting the extracted defect regions as connected defect regions by morphological processing;
Calculated for each connected defect region that connects the proportions of the contours along the exposure division side where the position in the image is preset in the contour of the connected defect region,
A defect classification method, wherein defects are classified based on the ratio. The defect classification method according to claim 6, wherein the extracted defect regions are connected by a morphological process using a structural element having a size larger than the width between adjacent chips. In a defect classification method for classifying defects in a subject that has been exposed to form a wiring pattern,
Taking an image of the subject surface after the exposure has been performed,
Extracting a defect area from the image,
Aggregating the extracted defect areas into subdivided exposure sections,
The adjacent defect sections of the defect sections aggregated with the defect areas are connected as connected defect sections,
Calculated for each connected defect section connecting the proportion of the contour along the exposure section side where the position in the image is preset among the outline of the connected defect section,
A defect classification method, wherein defects are classified based on the ratio. In a defect classification method for classifying defects in a subject that has been exposed to form a wiring pattern,
Taking an image of the subject surface after the exposure has been performed,
Inspecting the image in units of subdivisions of the exposure section, detecting a section where a defect exists as a defect section,
Connecting adjacent defect sections among the defect sections as connected defect sections;
Calculated for each connected defect section connecting the proportions of the contours along the exposure section side where the position in the image is preset among the contours of the connected defect sections,
A defect classification method, wherein defects are classified based on the ratio. 10. The defect classification method according to claim 8, wherein a section obtained by subdividing the exposure section is matched with a chip section.

Images