JP2008045887A - Inspection device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To realize accurate defect inspection by an inspection device for inspecting existence of a defect based on a subject image. <P>SOLUTION: In this inspection device for inspecting existence of a defect based on the subject 11 image, image acquisition means 1-8 acquire the subject 11 image, and a frame setting means 8 sets a frame in each prescribed pattern included in an image acquired by the image acquisition means, and a comparison means 8 compares two or more different frame images set by the frame setting means 8 in the state where each prescribed pattern included in each frame agrees with each other, and a defect detection means 8 detects a defect based on a comparison result by the comparison means. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an inspection apparatus that inspects for the presence or absence of a defect based on an image (video) of a subject such as a liquid crystal display (LCD) substrate, and more particularly to an inspection apparatus that realizes an accurate defect inspection.

For example, a system has been developed in which an image of an LCD substrate is captured by a camera and a defect on the surface of the LCD substrate is inspected based on the captured image (see, for example, Patent Document 1).
An example of a method for inspecting defects on the surface of the LCD substrate will be described with reference to FIGS.
FIG. 9 shows an example of an image of an LCD substrate (for example, a work substrate before cutting out individual LCD substrates). The symbols (A) to (I) and vertical and horizontal lines on the surface are described. For illustration purposes only, it is not an image of the LCD substrate.
In this example, the visual field of the screen (image) of the LCD substrate is divided into nine so that each divided area becomes a quadrangle slightly overlapping with the adjacent divided area. Specifically, each divided region is an equal rectangular region including the symbol parts (A) to (I).

FIG. 10A shows an example of an image of a certain divided area (surface A). In this divided area, there are pixel patterns 111 and 112, which are the original wiring patterns of the LCD substrate, and a defect 113, which appear in the image.
FIG. 10B shows an example of an image of a divided area (surface B) adjacent to the divided area shown in FIG. In this divided area, pixel patterns 121 and 122 which are the original wiring patterns of the LCD substrate exist, and these appear in the image.
Here, the original wiring pattern of the LCD substrate is ideally, for example, an equally spaced pattern, and the arrangement of the size, angle, and the like is the same for each pattern. Deviations occur in the pixel patterns 111, 112, 121, and 122 in the two divided regions.

FIG. 10C illustrates an example of a difference in luminance between the divided region (surface A) illustrated in FIG. 10A and the divided region (surface B) illustrated in FIG.
In this example, the screen (image) is simply divided equally, and the brightness of each surface is compared with other surrounding surfaces to determine whether or not there is a unique surface. As a result, the defect 113 can be detected, but the pixel patterns 111 and 121, which are the original wiring patterns, may be erroneously detected as defects.
The original wiring pattern is made of, for example, a metal such as chromium, or a transparent conductive film or insulating film.

JP 2004-117150 A

As described above, in the conventional defect inspection method, for example, as shown in FIGS. 10A, 10B, and 10C, in order to compare two adjacent divided regions (surface A and surface B). When the difference image is taken, a part of the pixel patterns 111 and 121 which are the original wiring patterns remain together with the pattern of the defect 113, and the part of the pixel patterns 111 and 121 which are the original wiring patterns are also peculiar. There has been a problem that it may be misrecognized as a point, and it has been required to improve the accuracy of defect inspection.
The present invention has been made in view of such a conventional situation, and an object of the present invention is to provide an inspection apparatus capable of realizing an accurate defect inspection.

In order to achieve the above object, the inspection apparatus according to the present invention inspects for the presence or absence of a defect based on an image of a subject with the following configuration.
That is, the image acquisition unit acquires an image of the subject. A frame setting unit sets a frame for each predetermined pattern included in the image acquired by the image acquisition unit. The comparison unit compares the images of two or more different frames set by the frame setting unit in a state where the images are adjusted so that the predetermined patterns included in the frames match each other. A defect detection means detects a defect based on the comparison result by the comparison means.
Therefore, since the images are compared in a state where the predetermined patterns included in the frames are matched, it is possible to realize a defect inspection with high accuracy.

Here, various objects may be used as the subject. For example, it is preferable to use an LCD substrate on which a plurality of identical (or almost identical) patterns are arranged.
In addition, as the image acquisition unit, for example, a unit that captures and acquires an image like a camera may be used, or a unit that inputs data of an already captured image from the outside may be used. .
Various patterns may be used as the predetermined pattern for setting the frame. For example, all or a part of the original wiring pattern, the pattern of the alignment mark, or the like can be used.
Moreover, as a frame, the thing of various magnitude | sizes and shapes may be used, for example, the square (or rectangle) larger than a predetermined pattern can be used.
In addition, as the frame setting, for example, a mode in which the frame is set once and used as it is may be used. Alternatively, after the initial setting, the frame is adjusted and set again (that is, after the adjustment) A mode in which a frame of () is set) may be used.

In addition, various numbers may be used as the number of two or more frames for comparing images. As an example, images of adjacent frames are compared.
In addition, the degree of coincidence in the case where adjustment is made so that the predetermined patterns included in each frame coincide with each other does not necessarily coincide completely, and for example, they coincide with each other to an extent that is practically effective.
In addition, various aspects may be used as an aspect for adjusting the predetermined patterns included in each frame to match, for example, the same arrangement relationship (for example, a relationship of size and angle) with respect to each predetermined pattern A mode in which the images in the frame are adjusted so that the frames set so as to match each other, or an image in the frame so as to match the predetermined pattern included in each frame set to include each predetermined pattern A mode of adjusting the angle can be used.
In such an adjustment, for example, a process of matching two or more predetermined patterns using a pattern matching process can be used.

  Various modes may be used as a mode for comparing two or more images to detect a defect. For example, by comparing the brightness of images, the difference in brightness becomes a predetermined threshold value or more ( Alternatively, a mode in which a portion exceeding a predetermined threshold value is regarded as a peak of a luminance difference and detected as a unique point (for example, a defective portion) can be used.

  As described above, according to the present invention, when inspecting for the presence or absence of a defect based on an image of a subject, a frame is set for each predetermined pattern included in the image of the subject, and the predetermined pattern included in each frame matches. Since the defect is detected by comparing the images of two or more different frames in the adjusted state, for example, it is prevented that the original wiring pattern is erroneously detected as a defect (or And accurate defect inspection can be realized.

Embodiments according to the present invention will be described with reference to the drawings.
FIG. 1 shows a configuration example of an inspection apparatus according to an embodiment of the present invention.
The inspection apparatus of this example includes an objective lens 1 having a role of a microscope for enlarging an image of an object, a CCD (Charge Coupled Device) camera 2 that constitutes an imaging apparatus that captures an image of the object, a light source unit 3, A light guide 4, an XY stage 5, a Z stage 6, a motor driver unit 7, a control unit 8 configured by, for example, a personal computer (PC) and controlling the entire apparatus, and a monitor 9 configuring a display device. I have. A subject 11 is placed on the XY stage 5. In this example, an LCD substrate (for example, a work substrate before cutting out individual LCD substrates) is used as the subject 11.
Various cameras (CCD camera 2 in this example) may be used. For example, an area sensor such as a television camera can be used.

An example of a schematic operation performed by the inspection apparatus of this example is shown.
An image of the subject 11 placed on the XY stage 5 is enlarged by the objective lens 1 and captured by the CCD element of the CCD camera 2, and an image signal (video signal) corresponding to the image of the subject 11 is sent from the CCD camera 2 to the control unit. 8 is transmitted.
At this time, the illumination light emitted from the light source unit 3 illuminates the subject 11 via the light guide 4. Further, the control unit 8 controls the motor driver unit 7 to move the XY stage 5 along the plane (XY plane), so that the position of the subject 11 placed on the XY stage 5 is a plane (XY plane). The objective lens attached to the Z stage 6 can be moved by moving the Z stage 6 in the vertical direction (Z direction) by controlling the motor driver unit 7. 1. The position of a set of the CCD camera 2, the light source unit 3, and the light guide 4 can be moved to an arbitrary position in the vertical direction (Z direction).

The control unit 8 displays an image of the subject 11 corresponding to the image signal received from the CCD camera 2 on the screen of the monitor 9.
Further, the control unit 8 has an image processing function, and by performing predetermined processing on the image of the subject 11, the appearance of the subject 11 is inspected, and the subject 11 (in this example, the LCD substrate) Detect defects that are present inside.
Here, in this example, a program describing a predetermined image processing procedure is stored in advance in the memory of the control unit 8, and the CPU (Central Processing Unit) of the control unit 8 reads and executes the program. A defect inspection process is executed.

A first embodiment of the present invention will be described.
In this example, an example of defect inspection processing performed by the inspection apparatus shown in FIG. 1 is shown.
The outline of the defect inspection process of this example will be described with reference to FIGS.
FIG. 2 shows an example of the image of the subject 11 acquired by the CCD camera 2 and displayed on the monitor 9, but the frame 31 is not an image of the subject 11 because it is set by the inspection apparatus.

In the subject 11 of the present example, the original wiring pattern of the LCD substrate is ideally arranged in the same pattern in the vertical and horizontal directions on the plane at equal intervals. Each wiring pattern is provided with a pattern (alignment mark) serving as a mark for alignment, for example. In this example, it is located at a fixed position (in this example, the upper left and lower right positions and the same position for each wiring pattern) at both ends of the square (or other shape) diagonal line on which each wiring pattern is provided. An alignment mark is provided.
In this example, the alignment mark is provided as a dedicated pattern different from the original wiring pattern. However, as another configuration example, a part of the original wiring pattern can be used as the alignment mark.

In the image of the subject 11 shown in FIG. 2, a plurality of pixel patterns 21 that are the original wiring patterns are arranged in the vertical direction and the horizontal direction. Here, for convenience of explanation, the upper left alignment mark 22 for each pixel pattern 21. The lower right alignment mark is not shown. A defect 23 generated in the subject 11 is also shown.
Here, each pixel pattern 21 is ideally the same pattern, but in reality, there is a shift in the arrangement of size, angle, and the like.

  In the image processing of this example, first, external alignment is performed, and each frame 31 is set (initially set) for each pixel pattern 21. In this example, the frame 31 is a square (or other shape) having a predetermined constant size (side length). As this fixed size, for example, a size is used so that the entire pixel pattern 21 can be included in each frame 31 even if there is a slight shift in the size of each pixel pattern 21.

  In the external alignment process, each frame 31 is arranged so that each pixel pattern 21 has the same arrangement with respect to the alignment mark 22 with reference to the upper left alignment mark 22 provided in each pixel pattern 21. Set. As an example, the distance (pitch) between the upper left alignment marks 22 in two different pixel patterns 21 adjacent in the vertical direction or the horizontal direction is detected, and each frame 31 is set based on the distance. It is possible. Here, the pattern shape of the alignment mark 22 is preset and stored in the memory of the control unit 8, and the control unit 8 analyzes the image of the subject 11 and performs, for example, pattern matching processing to thereby each pixel. The position of the alignment mark 22 in the pattern 21 is detected.

In the image processing of this example, next, internal alignment is performed, the size (side length) and angle of each frame 31 set for each pixel pattern 21 is adjusted, and the adjusted frame 31a is adjusted. Set.
FIG. 3A shows an example of a state before the arrangement of the frame 31 is adjusted. An upper left alignment mark 22 and an upper right alignment mark 41 are provided in the pixel pattern 21, and the size of the frame 31 set in the external alignment (in this example, a square having a side length a) is The angle of the frame 31 set in the external alignment is too large with respect to the actual pixel pattern 21 and is shifted with respect to the actual pixel pattern 21, and these are the detection positions of the two alignment marks 22 and 41. Is detected based on Note that the pattern shapes of the alignment marks 22 and 41 are preset and stored in the memory of the control unit 8, and the control unit 8 analyzes the image of the subject 11 and performs pattern matching processing, for example. The positions of the alignment marks 22 and 41 in the pixel pattern 21 are detected.

  FIG. 3B shows an example of a state after the arrangement of the frame 31 is adjusted. The arrangement of the pixel pattern 21 and the alignment marks 22 and 41 is the same as that shown in FIG. In adjusting the arrangement of the frame 31, the size and angle of the frame 31 are adjusted so as to correspond to the actual size and angle of the pixel pattern 21. In the adjusted frame 31a, the length b of the side is changed, and the angle is rotated by θ.

  Here, the adjusted frame 31 a does not necessarily match the pixel pattern 21. For example, the frame 31 a is a predetermined size of the pixel pattern 21 in advance and has a predetermined arrangement with respect to the pixel pattern 21. Conditions that determine the relationship are set, and the frame 31a is adjusted based on the conditions. As an example, the center point of the line segment connecting the positions of the two alignment marks 22 and 41 is matched with the center point of the frame 31a, and the line segment and the diagonal line of the frame 31a are parallel (in this example, they overlap). Thus, a condition can be used such that a result value obtained by multiplying the length of the line segment by a predetermined coefficient (for example, a value of 1 or more) becomes the length of the diagonal line of the frame 31a.

  The external alignment process and the internal alignment process may be performed using, for example, image data captured at the same magnification, or the external alignment process may be performed at a relatively low magnification. It is also possible to perform using the image data and then perform the internal positioning process using image data with a relatively high magnification.

Next, in the image processing of this example, after adjusting the arrangement of the frames 31 for each pixel pattern 21, the brightness of the adjacent pixel patterns 21 is compared.
As the positional relationship between the two pixel patterns 21 to be compared, for example, an arbitrary positional relationship such as up / down / left / right or diagonal may be used. A comparison with a number of other pixel patterns may be made.
In this example, two adjacent pixel patterns 21 are compared. However, as another configuration example, two non-adjacent pixel patterns 21 can be compared.
Here, since it is unclear which pixel pattern 21 is unique only by comparing two pixel patterns 21 and detecting only those that are specific to the luminance difference, The pattern 21 is compared with two or more other pixel patterns.

FIG. 4A shows an example of an image of a region (surface A) within a certain frame 31a. In this area, there are a pixel pattern 21 which is the original wiring pattern of the LCD substrate and two alignment marks 22 and 41, which appear in the image.
FIG. 4B shows an example of an image of a region (surface B) in another frame 61 adjacent to the region in the frame 31a shown in FIG. In this area, there are a pixel pattern 51 which is the original wiring pattern of the LCD substrate, two alignment marks 52 and 53, and a defect 23, which appear in the image.

  Here, in this example, since the size and angle of each frame 31a, 61 are adjusted so that the arrangement relationship with respect to each pixel pattern 21, 51 is equivalent, the two frames 31a, 61 to be compared are adjusted. If the sizes of the images in the frames 31a and 61 are adjusted so as to match (for example, one of them is enlarged or reduced as a whole), the pixel patterns 21 and 51 and the alignment marks 22 in the two frames 31a and 61 41, 52, and 53 also coincide with each other, and it is possible to detect only the defect 23 with high accuracy without being affected by these patterns.

FIG. 4C shows an example of the difference in luminance between the region (surface A) in the frame 31a shown in FIG. 4A and the region (surface B) in the frame 61 shown in FIG. 4B. Is shown. This difference in luminance is acquired by the control unit 8.
As shown in the drawing, for each surface, the luminance is compared with the other surrounding surfaces to determine whether there is a unique surface (for example, a surface where a luminance difference peak occurs). The surface where the defect 23 exists is detected as a specific surface because the brightness of the portion of the defect 23 is different from that of the other surface, and the presence of the defect 23 is detected.

With reference to FIG. 5, an example of the procedure of the defect inspection process performed by the inspection apparatus of this example is shown. This process is performed under the control of the control unit 8.
In the inspection apparatus, for example, when inspection is started in response to an instruction from the user, the XY stage 5 is first moved to, for example, specified coordinates in a state where the magnification of the objective lens 1 is set to the lowest possible magnification. In this way, the position of the subject 11 to be inspected is imaged by the CCD camera 2 (step S1). Next, the entire image on the screen is focused by AF (autofocus) (step S2). Thereafter, when the state for storing the captured image is set (step S3), the image data is stored in the memory (step S4), and then the image is registered as a processing target (step S5). On the other hand, when the state in which the captured image is not stored is set (step S3), the image is registered as a processing target without storing the image data (step S5). Then, it is determined whether or not the entire screen has sufficient luminance (for example, luminance equal to or higher than a preset threshold value) (step S6). If there is sufficient luminance, the inspection process is continued, and sufficient luminance is obtained. If there is no, the inspection process is terminated.

  If there is sufficient luminance on the entire screen, then, as shown in FIG. 2, for example, the screen is based on a pre-registered pixel pitch setting or at an arbitrary pixel pitch adapted to each pixel pattern 21. After setting each frame 31 in (image), as shown in FIGS. 3A and 3B, the position and angle of each frame 31 are adjusted, and thereafter, as shown in FIGS. As shown, the brightness of each of the surfaces of the frames 31a and 61 is compared with the surrounding surfaces (step S7). By such a luminance comparison, it is determined whether or not there is a peculiar surface (for example, a surface on which a luminance difference peak occurs) for each of the frames 31a and 61 (step S8), and there is no peculiar surface. If there is no defect, the inspection process is terminated.

  On the other hand, if there is a peculiar surface, it is determined whether or not the magnification of the objective lens 1 is sufficiently high (for example, a magnification greater than a predetermined threshold or the maximum possible magnification) (step S9) If the magnification is already sufficiently high, the inspection process is terminated. On the other hand, if the magnification is not yet sufficiently high, the XY stage 5 is moved so that the singular point of the detected singular surface (for example, the position where the peak of the luminance difference occurs) is located at the center of the screen, and the objective The lens 1 is switched to a lens with a higher magnification (for example, a magnification higher than the current value by a predetermined value) (step S10), and the same processing as in steps S2 to S10 is performed again, so that the image of the defect 23 is more detailed. Extract. In this example, since the original wiring pattern is a luminance comparison in the same pattern, only a singular point can be detected.

Here, in general, if the magnification of the objective lens 1 is too high, the luminance is too high and image processing may be difficult, or it may be difficult to detect a large defect. Inspection processing is performed by gradually changing the magnification to a higher magnification. Once measured at a low magnification and the approximate position of a peculiar part (in this example, a defective part) is grasped, when the measurement is switched to a higher magnification, the recognized defective part (or its part) It is possible to capture a unique portion in the screen by moving the coordinates corresponding to the vicinity) to be imaged. Furthermore, if the peculiar part is moved so as to be located at the center (or the vicinity thereof) of the screen and an image of that part is taken at a high magnification, it becomes very easy for the operator to see the defective part.
In addition, for example, an offset value of the start point of the detection area and the number of pixels (for example, the pixel size on the display screen) can be set so that the detection area in the screen (image) can be designated. May be.

In the pattern matching process, the following (Processing Example 1) to (Processing Example 3) can also be used.
(Processing Example 1) In order not to recognize a shake at the time of image acquisition or a manufacturing error of the substrate as a defect, a filter process ignoring ± n (n is an integer of 1 or more) pixels is performed to prevent erroneous recognition.
(Processing Example 2) In order not to recognize the luminance unevenness error at the time of manufacturing the substrate as a defect, filter processing is performed to ignore the luminance difference of ± v (v is a positive value) to prevent erroneous recognition.
(Processing Example 3) As correction of luminance unevenness of the illumination system, after subtracting the luminance unevenness from the acquired image, the frame 31 as shown in FIG. 2 is set.

  As described above, in the inspection apparatus of this example, when inspecting a defect in the screen (image) of the LCD substrate, each frame 31 is initially set for each pixel pattern 21, and then adjusted, Since each frame 31a according to the actual position and angle of the pixel pattern 21 is set, it is possible to compare the luminance in a state where the adjacent pixel patterns 21 are aligned. For this reason, for example, FIG. 9 and FIG. Compared with the detection method as shown in FIG. 5, it is possible to prevent the pixel pattern 21 which is the original wiring pattern from being detected as a singular point (defective part), and to prevent erroneous recognition. .

  Therefore, in the inspection apparatus of this example, for example, the defect detection accuracy in the pixel pattern 21 can be improved, and a defect can be detected by specifying a point without comparing all the fields of view. Here, in this example, in the defect inspection in the LCD substrate, the defect detection in the pixel pattern is mainly performed, and the pixel pattern in which the same pattern is arranged at an equal pitch is used. ing.

  In the inspection apparatus of this example, the control unit 8 controls the XY stage 5 and the Z stage 6 via the motor driver unit 7, and the object 11 is detected by the objective lens 1, the CCD camera 2, the light source unit 3, and the light guide 4. The image acquisition means is configured by the function of capturing and acquiring an image, and the control unit 8 sets a frame for the image of the subject 11 (in this example, it is set initially and then adjusted). The frame setting means is configured by the above, and the comparison means is configured by the function of comparing the luminance of the images of two or more frames in a state in which the control unit 8 has adjusted the pixel pattern which is the original wiring pattern to match. Thus, the defect detection means is configured by the function of the controller 8 detecting the defect based on the luminance comparison result.

A second embodiment of the present invention will be described.
In this example, an example of defect inspection processing performed by the inspection apparatus shown in FIG. 1 is shown.
With reference to FIG. 6 and FIG. 7, an outline of the defect inspection processing of this example will be described.
FIG. 6 shows an example of the image of the subject 11 acquired by the CCD camera 2 and displayed on the monitor 9, but the frame 81 is not an image of the subject 11 because it is set by the inspection apparatus.

In the subject 11 of the present example, the original wiring pattern of the LCD substrate is ideally arranged in the same pattern in the vertical and horizontal directions on the plane at equal intervals.
In the image of the subject 11 shown in FIG. 6, a plurality of pixel patterns 71 which are original wiring patterns are arranged in the vertical direction and the horizontal direction. In addition, a defect 72 generated in the subject 11 is also shown.
Here, each pixel pattern 71 is ideally the same pattern, but in reality, there is a deviation in the arrangement of size, angle, and the like.

In the image processing of this example, first, each frame 81 is set for each pixel pattern 71. In this example, as the frame 81, a square (or other shape) having a predetermined constant size (side length) is used. As this fixed size, for example, such a size that the entire pixel pattern 71 can be included in each frame 81 even if there is a slight shift in the size of each pixel pattern 71 or the like is used.
In this example, the field of view of the screen (image) is divided at a preset pixel pattern pitch, and each frame 81 is applied at equal intervals so that each pixel pattern 71 is included in each frame 81. In this example, each frame 81 is obtained by dividing the screen (image) at equal intervals, and includes each pixel pattern 71. Here, as an example of the pitch for dividing the screen (image), the same pitch as the pitch of the arrangement of the pixel patterns 71 can be used.

Next, in the image processing of this example, the luminance of the adjacent pixel patterns 71 is compared using the image in each frame 81 set for each pixel pattern 71.
As the positional relationship between the two pixel patterns 71 to be compared, for example, an arbitrary positional relationship such as up / down / left / right or diagonal may be used. A comparison with a number of other pixel patterns may be made.
In this example, two adjacent pixel patterns 71 are compared. However, as another configuration example, two non-adjacent pixel patterns 71 can be compared.
Here, normally, by comparing two pixel patterns 71 and detecting a peculiar difference in luminance, it is unclear which pixel pattern 71 has a peculiar one, so one pixel The pattern 71 is compared with two or more other pixel patterns.

FIG. 7A shows an example of an image of a region (surface A) within a certain frame 81. In this region, there is a pixel pattern 71 which is the original wiring pattern of the LCD substrate, and this appears in the image.
FIG. 7B shows an example of an image of a region (surface B) in another frame 101 adjacent to the region in the frame 81 shown in FIG. In this area, there are a pixel pattern 91 which is an original wiring pattern of the LCD substrate and a defect 72, which appear in the image.

  Here, in this example, when performing the luminance comparison, pattern matching processing is performed so that the two pixel patterns 71 and 91 included in the frames 81 and 101 match each other, and the two pixel patterns 71 and 91 to be compared are compared. The size and angle of the images in the frames 81 and 101 are adjusted so as to match (for example, one of them is enlarged or reduced as a whole and the other is rotated). Then, when performing the luminance comparison, the two pixel patterns 71 and 91 coincide with each other, and it is possible to detect only the defect 72 with high accuracy without being affected by these patterns.

  As described above, in this example, the internal patterns cut out by the frames 81 and 101 are subjected to pattern matching of the internal patterns between adjacent ones, so that the original pixel patterns 71 and 91 that are not defects are obtained. It adjusts so that it may correspond in luminance comparison. That is, even if divided at equal pitches, there is usually no deviation from the actual pixel pitch, so pattern matching processing is performed before the luminance comparison is performed for the surfaces of the frames 81 and 101. Thus, a slight shift for each divided surface is corrected.

  It should be noted that not all pixel patterns of the original wiring pattern are necessarily used as the pixel patterns 71 and 91 corresponding to the original wiring pattern to be subjected to such pattern matching. Some patterns effective as a pattern matching mark may be used, such as a pattern. As an example, it is also possible to use one or more alignment marks 22 and 41 as shown in FIG.

FIG. 7C shows an example of the luminance difference between the region (surface A) in the frame 81 shown in FIG. 7A and the region (surface B) in the frame 101 shown in FIG. Is shown. This difference in luminance is acquired by the control unit 8.
As shown in the drawing, for each surface, the luminance is compared with the other surrounding surfaces to determine whether there is a unique surface (for example, a surface where a luminance difference peak occurs). The surface where the defect 72 exists is detected as a specific surface because the brightness of the portion of the defect 72 is different from that of the other surface, and the presence of the defect 72 is detected.
In this example, the pattern matching process is performed in a state where the pixel patterns 71 and 91 are included in the frames 81 and 101, so that the offset patterns and the pixel patterns 71 and 91 in the frames 81 and 101 are matched. The brightness of two images is compared by adjusting the magnification and angle.

With reference to FIG. 8, an example of the procedure of the defect inspection process performed by the inspection apparatus of this example is shown. This process is performed under the control of the control unit 8.
In the inspection apparatus, for example, when inspection is started in response to an instruction from the user, the XY stage 5 is first moved to, for example, specified coordinates in a state where the magnification of the objective lens 1 is set to the lowest possible magnification. In this way, the position of the subject 11 to be inspected is imaged by the CCD camera 2 (step S11). Next, the entire image on the screen is focused by AF (autofocus) (step S12). After that, when the state for storing the captured image is set (step S13), after storing the image data in the memory (step S14), the image is registered as a processing target (step S15). On the other hand, when the state in which the captured image is not stored is set (step S13), the image is registered as a processing target without storing the image data (step S15). Then, it is determined whether or not there is sufficient luminance (for example, luminance equal to or higher than a preset threshold) on the entire screen (step S16). If there is sufficient luminance, the inspection process is continued and sufficient luminance is obtained. If there is no, the inspection process is terminated.

  If there is sufficient luminance on the entire screen, then, as shown in FIG. 6, after setting each frame 81 on the screen (image) based on, for example, a preset pixel pitch setting, Matching processing is performed, and as shown in FIGS. 7A to 7C, the brightness of each of the surfaces of the frames 81 and 101 is compared with the surrounding surfaces (step S17). By such a luminance comparison, it is determined whether or not there is a peculiar surface (for example, a surface where a luminance difference peak has occurred) for each of the frames 81 and 101 (step S18). If there is no defect, the inspection process is terminated.

  On the other hand, if there is a peculiar surface, it is determined whether or not the magnification of the objective lens 1 is sufficiently high (for example, a magnification greater than a predetermined threshold or the maximum possible magnification) (step S19) If the magnification is already sufficiently high, the inspection process is terminated. On the other hand, if the magnification is not yet sufficiently high, the XY stage 5 is moved so that the singular point of the detected singular surface (for example, the position where the peak of the luminance difference occurs) is located at the center of the screen, and the objective The lens 1 is switched to a lens with a higher magnification (for example, a magnification higher than the current value by a predetermined value) (step S20), and the same process as in steps S12 to S20 is performed again, so that the image of the defect 72 is more detailed. Extract. In this example, since the original wiring pattern is a luminance comparison in the same pattern, only a singular point can be detected.

Here, in general, if the magnification of the objective lens 1 is too high, the luminance is too high and image processing may be difficult, or it may be difficult to detect a large defect. Inspection processing is performed by gradually changing the magnification to a higher magnification. Once measured at a low magnification and the approximate position of a peculiar part (in this example, a defective part) is grasped, when the measurement is switched to a higher magnification, the recognized defective part (or its part) It is possible to capture a unique portion in the screen by moving the coordinates corresponding to the vicinity) to be imaged. Furthermore, if the peculiar part is moved so as to be located at the center (or the vicinity thereof) of the screen and an image of that part is taken at a high magnification, it becomes very easy for the operator to see the defective part.
In addition, for example, an offset value of the start point of the detection area and the number of pixels (for example, the pixel size on the display screen) can be set so that the detection area in the screen (image) can be designated. May be.

In the pattern matching process, the following (Processing Example 1) to (Processing Example 3) can also be used.
(Processing Example 1) In order not to recognize a shake at the time of image acquisition or a manufacturing error of a substrate as a defect, a filter process that ignores ± n (n is an integer of 1 or more) pixels is performed to prevent erroneous recognition.
(Processing Example 2) In order not to recognize the luminance unevenness error at the time of manufacturing the substrate as a defect, filter processing is performed to ignore the luminance difference of ± v (v is a positive value) to prevent erroneous recognition.
(Processing Example 3) As correction of luminance unevenness of the illumination system, after subtracting luminance unevenness from the acquired image, a frame 81 as shown in FIG. 6 is set.

  As described above, in the inspection apparatus of this example, when inspecting a defect in the screen (image) of the LCD substrate, each frame 81 is set for each pixel pattern 71 and pattern matching processing is performed. 81. Since the luminance comparison is performed by adjusting the images so that the pixel patterns 71 included in 81 match each other, the luminance can be compared in a state where the adjacent pixel patterns 71 are combined. 9 and FIG. 10 can prevent the pixel pattern 71, which is the original wiring pattern, from being detected as a singular point (defective part) and prevent erroneous recognition. can do.

  Therefore, in the inspection apparatus of this example, for example, the defect detection accuracy in the pixel pattern 71 can be improved, and a defect can be detected by point designation without comparing all the fields of view. Here, in this example, in the defect inspection in the LCD substrate, the defect detection in the pixel pattern is mainly performed, and the pixel pattern in which the same pattern is arranged at an equal pitch is used. ing.

  In the inspection apparatus of this example, the control unit 8 controls the XY stage 5 and the Z stage 6 via the motor driver unit 7, and the object 11 is detected by the objective lens 1, the CCD camera 2, the light source unit 3, and the light guide 4. An image acquisition unit is configured by a function of capturing and acquiring an image, and a frame setting unit is configured by a function of the control unit 8 setting a frame with respect to the image of the subject 11. Comparing means is configured by the function of comparing the brightness of the images of two or more frames in a state adjusted by the pattern matching process so that the pixel pattern as the wiring pattern matches, and the control unit 8 determines the brightness comparison result. Defect detecting means is constituted by the function of detecting defects based on the above.

Here, the configuration of the system and apparatus according to the present invention is not necessarily limited to the configuration described above, and various configurations may be used. The present invention can also be provided as, for example, a method or method for executing the processing according to the present invention, a program for realizing such a method or method, or a recording medium for recording the program. It is also possible to provide various systems and devices.
The application field of the present invention is not necessarily limited to the above-described fields, and the present invention can be applied to various fields.
In addition, as various processes performed in the system and apparatus according to the present invention, for example, the processor executes a control program stored in a ROM (Read Only Memory) in hardware resources including a processor and a memory. A controlled configuration may be used, and for example, each functional unit for executing the processing may be configured as an independent hardware circuit.
The present invention can also be understood as a computer-readable recording medium such as a floppy (registered trademark) disk or a CD (Compact Disc) -ROM storing the control program, and the program (itself). The processing according to the present invention can be performed by inputting the program from the recording medium to the computer and causing the processor to execute the program.

It is a figure which shows the structural example of the test | inspection apparatus which concerns on one Example of this invention. It is a figure which shows an example of the setting of the frame which concerns on 1st Example of this invention. It is a figure which shows an example of adjustment of the frame which concerns on 1st Example of this invention. It is a figure which shows an example of the comparison of the brightness | luminance which concerns on 1st Example of this invention. It is a figure which shows an example of the flow of the defect inspection which concerns on 1st Example of this invention. It is a figure which shows an example of the setting of the frame which concerns on 2nd Example of this invention. It is a figure which shows an example of the comparison of the brightness | luminance which concerns on 2nd Example of this invention. It is a figure which shows an example of the flow of the defect inspection which concerns on 2nd Example of this invention. It is a figure which shows an example of the setting of a frame. It is a figure which shows an example of a brightness | luminance comparison.

Explanation of symbols

  1 .... objective lens, 2 .... CCD camera, 3 .... light source unit, 4 .... light guide, 5 .... XY stage, 6 .... Z stage, 7 .... motor driver unit, 8 .... control unit, 9 ..Monitor, 11..Subject, 21, 51, 71, 91, 111, 112, 121, 122..Pixel pattern, 22, 41, 52, 53..Alignment mark, 23, 72, 113..Defective , 31, 31a, 61, 81, 101 .. frame,

Claims (1)

In an inspection apparatus for inspecting the presence or absence of defects based on an image of a subject,
Image acquisition means for acquiring an image of the subject;
Frame setting means for setting a frame for each predetermined pattern included in the image acquired by the image acquisition means;
Comparison means for comparing two or more different frame images set by the frame setting unit in a state where the predetermined patterns included in each frame are adjusted so as to match,
A defect detection means for detecting a defect based on a comparison result by the comparison means;
An inspection apparatus comprising:

Images