JP3829024B2 - Display screen inspection method and apparatus - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a display screen inspection method and apparatus, and more particularly, a display screen inspection for inspecting defects in display characteristics due to a large number of display pixels constituting a display screen such as a cathode ray tube (CRT), a liquid crystal display panel, a plasma display panel and the like. It relates to a method and a device.
[0002]
[Prior art]
The inspection of the display screen is usually performed by visual determination by a person or automatic determination with image processing by an automatic machine. In the case of humans, even if the model to be inspected changes, it can be easily handled and the rise is quick. However, it takes a long time to specify the detailed position of a defective pixel, resulting in a decrease in throughput. In addition, there is a problem in maintaining and unifying inspection accuracy due to individual differences in inspection. In the case of using an automatic machine, the above-mentioned problem caused by a person can be solved. However, when the model to be inspected is switched, it takes a lot of time for adjustment such as setting of conditions for the inspection.
[0003]
Inspection by an automatic machine is particularly effective for the above-mentioned CRT, liquid crystal display panel, plasma display panel and the like having many display pixels, and has become mainstream. In addition, with the recent increase in the size of display screens and the increase in the density of display pixels, the number of pixels has increased and the speed of inspection has been further increased.
[0004]
In order to cope with this, a method of compressing and inspecting the size of the display screen is conventionally used. If the screen size spread in two dimensions can be compressed to ½ in both the vertical and horizontal directions, the total number of pixels is reduced to ¼, and the subsequent processing speed is quadrupled. Such size reduction of the display screen is performed by, for example, four each approaching in a predetermined section a with respect to the luminance characteristic data shown in FIG. 8A detected for each pixel constituting the display screen. Each set of four data corresponding to a pixel is compressed into one data b1, for example, taking an average value as shown in (a1) of FIG. 8, and each compressed data b1 compressed thereafter is used as reference data. In comparison, the compressed pixel defect is determined and detected.
[0005]
Specifically, the four data 99, 100, 100, 100 at the vertical and horizontal addresses 0 · 0, 0 · 1, 1, · 0, 1 · 1 shown in FIG. The data is compressed into one data 100 indicated by vertical and horizontal addresses 1 and 1 in FIG.
[0006]
[Problems to be solved by the invention]
However, when the inspection is performed by the compression method as described above, there is a case where an inspection error occurs and a defective pixel cannot be detected. In the unlikely case, it may lead to a decrease in yield and reliability after the final product is obtained. .
[0007]
Accordingly, the present inventors have conducted various experiments and studies on this, and found that there is a problem with the conventional compression method. More specifically, the combination of data to be compressed as in the conventional case is compared with the designated position of the section a in the case shown in FIG. 8A, for example, in (b) to (d) of FIG. If the positions a1 to a3 are specified with slightly different positions, the data corresponding to the defective portion of the data of each pixel does not remain on the compressed data, and the defect may not be detected.
[0008]
The values of the pixel data handled in FIGS. 8A to 8D and the arrangement positions thereof are common to each other. For example, if the luminance of 120 or more is excessive and defective, four adjacent pixels All have defect image portions c whose data are 146, 151, 154, 142. Therefore, when these are compared with the reference data, they are determined as defective pixels and detected.
[0009]
However, the four pieces of pixel data 146, 151, 154, 142 of the defective image portion c are shown in FIG. 8 depending on the position of the division designation for each compression operation as shown in (a) to (d) of FIG. 8 (a) is located at vertical and horizontal addresses 5, 6, 5, 7, 6, 6, 6 and 7, and belongs to two upper and lower data compression sections a. ) Are located at the vertical and horizontal addresses 4, 5, 4, 6, 5, 5, 5 and 6, and belong to the two left and right data compression sections a1., And in FIG. 8C, the vertical and horizontal addresses. 5 · 5 · 5 · 6 · 6 · 5 · 6 · 6 and belongs to four combination sections a2,... Vertically and horizontally, and in FIG. 6, 4, 7, 5, 6, 5 and 7, all four belong to one data compression section a3.
[0010]
The upper pixel data 146 and 151 of the defective image portion c in FIG. 8A are averaged in combination with the other two pixel data 102 and 101 in the same section a, and the vertical and horizontal addresses in FIG. 8A1 are taken. The lower pixel data 154 and 142 which are compressed into one data 125 at 3 and 4 and the defective image portion c remains are averaged in combination with the other two pixel data 106 and 100 in the same section a. 8 is compressed into one data 126 at vertical and horizontal addresses 4 and 4 in (a1), and any compressed pixel in any section a to which they belong is determined as a defect in comparison with the reference data 120 and detected. The defect position is specified from the address data of the detected position.
[0011]
The left pixel data 146 and 154 of the defective image portion c in FIG. 8B is averaged in combination with the other two pixel data 100 and 100 in the same section a1, and the vertical and horizontal addresses in FIG. 8B1 are taken. The right pixel data 151 and 142 which are compressed into one data 125 in 3 and 3 and the defect image portion c is left are averaged in combination with the other two pixel data 102 and 100 in the same section a1 as shown in FIG. Compressed into one data 124 at the vertical and horizontal addresses 3 and 4 in (b1) of (b1), any compressed pixel of the section a1 to which they belong is determined as a defect in comparison with the reference data 120 and detected. The defect position is identified from the address data of the determined position.
[0012]
The upper left pixel data 146 of the defective pixel c in FIG. 8C is averaged in combination with the other three pixel data 105, 102, 100 in the same section a2, and the vertical and horizontal addresses 3 in FIG. 8C1 are taken. 8 is compressed into one data 113 at 3, and the upper right pixel data 151 of the defective pixel c is averaged in combination with the other three pixel data 101, 100, and 102 in the same section a2 (c1 in FIG. ) Is compressed into one data 114 at the vertical and horizontal addresses 3 and 4, and the lower left pixel data 154 of the defective pixel c is averaged in combination with the other three pixel data 100, 108, 106 in the same section a2. The lower right pixel data 142 of the defect image c is compressed into one data 117 at the vertical and horizontal addresses 4 and 3 in FIG. 8C1, and the other three pixel data 10 in the same section a2. , 100 and 100 are averaged and compressed into one data 111 at the vertical and horizontal addresses 4 and 4 in (c1) of FIG. 8, and the compressed data of any section a2 to which they belong belongs to the reference data 120 Are not determined as defects and are not detected.
[0013]
The four pixel data 146, 151, 154, 142 on the upper, lower, left, and right sides of the defective pixel c in FIG. 8D are averaged as they are in the same section a3, and the vertical and horizontal addresses 4 in FIG. 8D1 are taken. Compressed into one data 148 at 4, and detected as a defect in comparison with the reference data 120 at the compressed pixel in the section a3 to which they belong, and the defect position is determined from the address data of the detected position Identify.
[0014]
This means that even if the data arrangement is the same, the detection result of the defective portion is greatly different depending on the designated position of the section where the compression operation for detecting the defect is performed. The data of the defect image c is included in one designated section. Even if it exists, unless all the data in the one section are defective, it may happen that the compressed data does not remain as defect information due to the difference in the combination of the values of a plurality of data to be compressed. Since the position, shape, and size at which a defect occurs in each display pixel in manufacturing the display device is not constant, the above-described conventional device cannot avoid this problem. When you want to obtain a more accurate inspection result by repeatedly inspecting the same display screen, it is also possible to eliminate an imaging position shift of about one pixel due to each imaging, resulting in mixed inspection results To do. If the number of data to be compressed is halved, the probability of occurrence of the problem is halved, but the problem is unavoidable and the processing speed is greatly reduced.
[0015]
An object of the present invention is to provide a display screen inspection method and apparatus capable of performing defect inspection with high reliability and being capable of performing defect inspection with high reliability while allowing high-speed processing at a high compression rate.
[0016]
[Means for Solving the Problems]
In order to achieve the above object, a display screen inspection method according to the present invention is a method corresponding to a plurality of pixels included in a predetermined section, with respect to data of a predetermined display characteristic detected for each pixel constituting the display screen. Each set of data is compressed into one data, each compressed data is handled as compressed image data having an array following each corresponding compression pixel position, and this compressed image data is a predetermined compression target. Compressed pixels corresponding to the compressed image data under each shift condition are obtained for each shift condition shifted to a plurality of positions satisfying the condition that some pixels of the same segment overlap and exist in the designated position of the segment Mainly, the data is combined into one determination target compressed data, and each determination target compressed data is compared with reference data to determine and detect a defect in display characteristics for each compression pixel. It is an butterfly.
[0017]
As a display screen inspection apparatus that achieves this, detection means for detecting predetermined display characteristics for each pixel constituting the display screen, and each detected data can be handled with vertical and horizontal addresses corresponding to the arrangement of each pixel. Each of the pixel data stored in the storage means and a set corresponding to a plurality of pixels included in the predetermined section, respectively, is compressed into one data, and each compressed data is compressed. A condition in which compressed image data having an arrangement following each compressed pixel position corresponding to them is used, and this compressed image data has a specified position of the predetermined section to be compressed and a part of pixels in the same section overlap. The image compression means for generating each shift condition that has been shifted to a plurality of positions satisfying the above, and the compressed pixel data corresponding to the compressed image data under each shift condition are combined into one determination target pressure. Compressed image combining means for combining with data, and defect determining means for comparing each of the combined determination target compressed data with reference data to determine and detect a display characteristic defect for each compressed pixel. Features are enough.
[0018]
In these configurations, in order to detect defective pixels by comparing pixel data having predetermined characteristics of each pixel of the display screen with reference data, a plurality of each of the pixel data is compressed into one compressed data. Each compressed data is treated as compressed image data having an array following each corresponding compressed pixel position, and this compressed image data is assigned to a specified position of a predetermined section to be compressed by some pixels in the same section. Obtaining for each shift condition that has been shifted to a plurality of positions while satisfying the overlapping conditions, and combining the corresponding compressed pixel data of the compressed image data under each shift condition into one determination target compressed data Thus, the obtained determination target compressed data is a combination of surrounding pixel data per pixel data according to a shift condition in which some pixel data in the same section overlap. Since it is easy to pick up the data of the defective image portion that has been easily lost due to the difference in the combination of a plurality of pixel data to be compressed, the determination target obtained in this way By comparing the compressed data with the reference data, it is possible to determine and detect defects with high accuracy without performing any particular processing time by simply performing the shift processing and the composition processing. The defect position is specified from the address data of the position.
[0019]
The predetermined display characteristics can be inspected for various types such as brightness and color characteristics. For the combination of a plurality of compressed pixel data, an average value of corresponding compressed image data may be obtained, or one of a median value, a maximum value, and a minimum value may be obtained. The maximum value is particularly suitable for detecting over-brightness defects because it does not miss over-brightness pixels, and the minimum value is particularly suitable for detection of under-brightness defects because it does not miss under-brightness pixels. This is effective when detecting a defect having an area in which there are a plurality of defects in a specified category.
[0020]
Further objects and features of the present invention will become apparent from the following detailed description and drawings. Each feature of the present invention can be used alone or in combination in various combinations as much as possible.
[0021]
DETAILED DESCRIPTION OF THE INVENTION
In the following, embodiments of the display screen inspection method and apparatus of the present invention will be described with reference to FIGS. 1 to 7 together with the examples thereof for the understanding of the present invention.
[0022]
This embodiment is an example of a case where a liquid crystal display (LCD) panel is used as an object to be inspected and an inspection for detecting an excessive luminance pixel of each pixel as a defect is performed. However, the present invention is not limited to this, and is effective when applied to inspection of display screens of various display devices having display screens composed of a large number of display pixels, including CRTs and plasma display panels. In addition, the display characteristics to be defective are not limited to the above excessive luminance, and defects such as insufficient luminance or their color characteristics can be detected in the display pixels of the color display screen.
[0023]
The display screen inspection method of the present embodiment is performed using, for example, an inspection apparatus as shown in FIG. 1, and the inspection object 1 is placed on the inspection stage 2 and is higher than the inspection origin on the apparatus. Accurate positioning is performed. This is useful, for example, for recognizing the position of each pixel constituting the display screen 1a of the device under test 1 with high accuracy. The display screen 1a of the inspected object 1 is also imaged from above by the imaging camera 3 that is positioned with high accuracy with respect to the inspection origin on the apparatus. The imaging camera 3 uses, for example, a CCD area sensor 3a. The CCD area sensor 3a has sensor pixels 3a1 that correspond one-to-one with the pixels 1a1 arranged in a wide and vertical direction on the display screen 1a. The display state of each pixel 1a1 on the display screen 1a of the inspection object 1, that is, the driving state is received by each sensor pixel 3a1 and output to the outside as an electrical signal. In the present embodiment, a luminance signal may be used, so that monochrome density detection data can be used as it is. However, the CCD area sensor 3a may be of a size that can collectively capture the entire integral of the display screen 1a of the object 1 to be inspected, but may capture the entire screen while sequentially shifting the partial screen.
[0024]
The imaging information of the display state at each pixel 1a1 of the display screen 1a captured by the imaging camera 3 is input to the microcomputer 4 and stored in the image memory 5. Here, image information from each pixel 1a1 is stored in the image memory 5 under the address designation corresponding to the arrangement position of each pixel 1a1 on the display screen 1a, and the image information for each pixel 1a1 is stored in each pixel 1a1. It can be handled according to the actual array position. However, the specific data development method can be freely selected. In addition, it is convenient for later use that the data stored in the image memory 5 is related to the display characteristics to be inspected in the imaging information. In the present embodiment, the display luminance data is for each pixel 1a1. . 3A to 3D, the vertical and horizontal addresses are specified and stored so that each pixel data 1a2 stored in the image memory 5 can be handled corresponding to the actual arrangement of each pixel 1a1 on the display screen 1a. The state which was made is shown typically.
[0025]
As described above, when predetermined imaging data is stored in the image memory 5 in a predetermined format, an image processing algorithm included in or received from others according to the program stored in the program memory 6 by the microcomputer 4 Is used to perform inspection processing for determining and detecting defects for each pixel data 1a2 of the display screen 1a imaged from the above-described imaging data stored in the image memory 5.
[0026]
In the case of this embodiment, this inspection is performed for a plurality of pixel data 1a2 having a predetermined display characteristic detected for each pixel 1a1 constituting the display screen 1a, for example, included in a predetermined section S shown in FIG. Each set of pixel data 1a2 corresponding to a pixel is compressed into one compressed data 1a3 shown in (a1) of FIG.
[0027]
For example, taking FIG. 3A as an example, the combinations of the pixel data 1a2 in the section S including the vertical and horizontal addresses 0 · 0, 0 · 1, 1, · 0, 1 · 1 are 99, 100, 100, 100. The result is compressed as 100 at the vertical and horizontal addresses 1 and 1 in FIG. Similar image compression is performed on the other pixels 1a1 under the designation of the division S as shown in FIG.
[0028]
Next, the compressed data 1a3 compressed as described above is handled as compressed image data D1 as shown in FIG. 3A1 in an array following the compressed pixel positions corresponding to the compressed data 1a3, and becomes the compression target. As shown in (b), (c), and (d) of FIG. 3, satisfying the condition that a part of the pixel data 1a2 in the same section S exists at the designated position of the predetermined section S. The compressed image data D2 to D4 for each shift condition shifted to a plurality of positions are obtained as shown in (b1) to (d1) of FIG. In FIG. 3B, the designated position of the section S is shifted by one pixel vertically and horizontally with respect to the case of FIG. 3A. For example, the pixel data 100 at the vertical and horizontal addresses 1 and 1 is shown in FIG. ) And (b) are duplicated in both sections S, and are used for the compression in combination with other data different in each section S. In FIG. 3C, the designated section S is shifted by one pixel in the horizontal direction with respect to the case of FIG. 3A. For example, the pixel data 100, 100 of the vertical and horizontal addresses 0, 1, 1, 1. 3 are duplicated in both sections S of FIGS. 3A and 3C, and are used for the compression in combination with other data different in each section S. In FIG. 3D, the designated section S is shifted by one pixel in the vertical direction with respect to the case of FIG. 3A. For example, the pixel data 100, 100 of the vertical and horizontal addresses 1 · 0, 1 · 1 are 3 are duplicated in both sections S of (a) and (d) of FIG. 3, and are used for the compression in combination with other data different in each section S. As described above, the segment position shown in (a) of FIG. 3 is obtained by performing the compression operation under each shift condition by shifting the segment S as the data compression target shown in (a) to (d) of FIG. Then, compressed image data D1 is obtained as shown in (a1) of FIG. 3, and compressed image data D2 is obtained as shown in (b1) of FIG. ), The compressed image data D3 is obtained as shown in (c1) of FIG. 3, and the compressed image data D4 is obtained as shown in (d1) of FIG. 3 at the divided position of FIG.
[0029]
In the embodiment shown in FIG. 3, the pixel data handled in (a) to (d) of FIG. 8 and the arrangement thereof are set in the same manner as the position setting of each designated segment. It is easy to compare with the conventional method which performs only one compression operation described in the above.
[0030]
Next, the compressed data 1a3 corresponding to each other in, for example, the same vertical and horizontal addresses of the compressed image data D1 to D4 shown in (a1), (b1), (c1), and (d1) of FIG. Are combined with one determination target compressed data 1a4 as shown in (e), and each determination target compressed data 1a4 is compared with reference data, for example, 120 to determine and detect a display characteristic defect for each compression pixel. The determination target compressed data 1a4 is, for example, 100, 99, 100, and 99 for the compressed pixel data 1a3 at the vertical and horizontal addresses 1 and 1 of (a1), (b1), (c1), and (d1) in FIG. Therefore, it is synthesized into one data 99 at the vertical and horizontal addresses 1 and 1 shown in FIG. The defect image E represented by the data of the vertical and horizontal addresses 3 and 4 can be determined and detected from the array of composite image data of the respective determination target compressed data 1a4 as shown in FIG. The position of the defect image E is specified from the detected vertical and horizontal addresses 3 and 4. As a result, it is needless to say that it is not affected by a shift of about one pixel level of the segment position with respect to the captured pixel data considered due to a position shift between the CCD area sensor 3a and the inspection object 1. Further, the interlocking of the compression by averaging the plurality of pixel data 1a2 as described above and the synthesis processing of the plurality of compressed image data D1 to D4 obtained thereby is a state in which the display characteristics of several pixels 1a1 are leveled. This is especially effective when you want to compare with the reference level.
[0031]
In the above embodiment, the compressed image data D1 to D4 are combined by taking an average value. As another example of the synthesis, a synthesis process for obtaining a median value, a maximum value, a minimum value, or the like may be performed, and the maximum value is suitable for detection of excessive luminance because an excessive luminance defect is not missed. The minimum value is suitable for detecting defects with insufficient brightness because it does not miss defects with insufficient brightness, and is effective when detecting defects with an area that has more than one defect in the specified category. It is.
[0032]
In the inspection method as described above, as shown in FIG. 1, the imaging information from the imaging camera 3 serving as detection means for detecting a predetermined display characteristic for each pixel constituting the display screen 1a of the device under test 1 is micro-measured. As shown in FIG. 2, the microcomputer 4 receives the imaging information from the imaging camera 3 and stores the display characteristic data to be inspected from the imaging information as an image memory. 5 is determined based on the input unit 21 to be stored under the designation of the address corresponding to the two-dimensional array of display pixels 1a1 and the pixel data 1a2 stored in the image memory 5. In order to detect and detect the position, each pixel data 1a2 stored in the image memory 5 corresponds to a plurality of pixels included in a predetermined section. Each set is compressed as a piece of compressed data 1a3, and each compressed data 1a3 is used as compressed image data having an array corresponding to each compressed pixel position corresponding to the compressed data 1a3. Image compression that is generated as compressed image data D1 to D4 for each shift condition in which the specified position of each section is shifted to a plurality of positions while satisfying the condition that some pixels 1a1 in the same section S overlap. An image compression unit 22 as a means, and a compressed image composition unit 23 as a compressed image composition means for synthesizing compressed pixel data corresponding to compressed image data in each shift condition into one determination target compressed data. In addition, the defect determination unit 24 as a defect determination unit that detects the display characteristic defect for each compression pixel by comparing each determination target compressed data with reference data. The provided, it may be sequentially Hatarakasere accordance with the process of rank as these to a flowchart of steps # 1 to step # 4 shown in FIG. However, these processes can be performed not only by the microcomputer 4 but also by a dedicated processing circuit and processing apparatus.
[0033]
It is conceivable that the image compression unit 22 does not need to generate all shift-compressed images. For example, when the captured image F of the elongated inspected object 1 as shown in FIG. 4 is an inspection target, the relative displacement in the lateral direction with an extremely small number of data is important for detecting the defective portion E without missing it. Teku The When compression is performed at 1/5 horizontal and 1/5 horizontal, 25 types of compressed images can be considered. However, as shown in FIG. Thus, the processing speed can be increased without affecting the inspection accuracy. It should be noted that it is desirable to change the number of generated shift-compressed images according to the shape of the defective portion E to be detected and the direction of the relative displacement that can occur.
[0034]
Furthermore, regarding smoothing compression that takes the average value of the target data, it may be possible to set the filter elements in advance as shown in FIG. FIG. 6A shows a filter element FE1 of 1/2 horizontal and 1/2 vertical, and four pixel data of 2 × 2 horizontal in one section S to be compressed are shifted to four positions. In response to the compression and synthesis, the number of times each of the 9 pixel data of 3 × 3 pixels including the surrounding pixel data handled at each shift position is added is calculated for each pixel. The center pixel is the maximum number of times 4. FIG. 6B shows a filter element FE2 having a width of 1/3 and a height of 1/3, and nine pixel data of 3 × 3 in one section S to be compressed are shifted to nine positions. In response to the compression and synthesis, the number of times each of the 25 pixel data of 5 × 5 pixels including the surrounding pixel data handled at each shift position is added. It is shown corresponding to the pixel, and the center pixel is the maximum number of times 9. FIG. 6C shows a filter element FE3 having a width of 1/4 and a length of 1/4, and 16 pieces of pixel data arranged in 4 × 4 in one section S to be compressed have 16 positions. The number of times that each of 49 pixel data of 7 × vertical × 7 including pixel data handled at each shift position is subjected to addition processing in response to the shift and compression to synthesize. Is shown corresponding to each pixel, and the maximum number of times is 16 in the center pixel.
[0035]
The specific usage of the 1/2 horizontal and 1/2 vertical filter elements in FIG. 6A will be described. The horizontal 2 × vertical 2 indicated by the hatched portions in FIGS. 7A to 7D. In the case of compressing the partial image 31 of the pixels of FIG. 6, in relation to the shift in the horizontal and vertical directions, a partial image 32 of 9 pixels of 3 × 3 inclusive is included in FIG. Apply the filter element. FIG. 7 (e) shows this by generalizing each pixel data by adding address codes a11 to a13, a21 to a23, a31 to a33, and each pixel data a11 to a13, a21 to a23, Each value of a31 to a33 is multiplied by the number of times of addition at the corresponding pixel position of the filter element FE1 shown in FIG. 6A, and the value obtained by adding the multiplication results is 16 pixel data. When division is performed, data is obtained by smoothing, compressing and synthesizing as shifts at four positions. This can be output as a compressed density value when calculated by the microcomputer 4 as follows.
[0036]
(1/16) × [(a11 + a13 + a31 + a33) + 2 × (a12 + a21 + a23 + a32) + 4 × a21]
[0037]
【The invention's effect】
According to the present invention, as is apparent from the above description, the combination of the surrounding pixel data with respect to one pixel data is made different by the number of shifts according to the shift condition in which a part of the pixel data of the same section overlaps. A plurality of pieces of pixel data to be compressed by obtaining the compression data to be determined including the compressed information, comparing it with the reference data, determining and detecting defects in the display characteristics of the display screen, and specifying their positions It is easier to pick up the data of defective image parts that have been easy to be missing due to the difference in combination, and it is possible to detect and detect defects with high accuracy without performing processing time by simply performing the shift process and the composition process. The inspection accuracy can be improved without impeding the speeding up of the process.
[Brief description of the drawings]
FIG. 1 is a schematic configuration diagram of a display screen inspection apparatus according to an embodiment of the present invention.
FIG. 2 is a flowchart showing main processing units that perform inspection processing of the apparatus of FIG. 1 in accordance with the flow of processing work;
FIGS. 3A and 3B are explanatory views schematically showing one specific embodiment operation in the image compression unit and the compressed image synthesis unit in FIG. 1, wherein (a) to (d) are set for a pixel row to be inspected; Each of the shift positions of the compression target section is shown, (a1) to (d1) are examples of compressed images at the respective shift positions, and (e) is a compressed composite image obtained by combining the compressed images.
FIG. 4 is a diagram illustrating an example of a captured image when a vertically long display screen is imaged and inspected.
5 is an explanatory diagram showing another compression method suitable for the captured image of FIG. 5. FIG.
FIG. 6 is a schematic diagram illustrating an example of a filter element used for image compression and composition operations.
7 is an explanatory diagram when compressing and synthesizing an image using the filter element of FIG. 6, in which (a) to (d) show different positions of compression sections, and (e) shows one compression. It is a partial image figure which generalizes and shows the partial image containing the process target pixel by shift operation about an object division.
FIG. 8 shows an example of an image compression operation in a conventional inspection method, in which (a) to (d) show positions of compression object sections set differently for the same inspection object pixel column, and (a1) ˜ (d1) shows examples of compressed images at each setting position.
[Explanation of symbols]
1 Inspected object
1a Display screen
1a1 Display pixel
1a2 Pixel data
1a3 compressed data
1a4 Determination target compressed data
3 Imaging camera
3a CCD area sensor
3a1 Sensor pixel
4 Microcomputer
5 Image memory
6 Program memory
21 Image input section
22 Image compression unit
23 Compressed image composition unit
24 Defect determination unit
D1-D4 compressed image data

Claims (7)

  A set of data corresponding to a plurality of pixels included in a predetermined section is compressed into one data for each data of a predetermined display characteristic detected for each pixel constituting the display screen, and each compressed compressed data Are treated as compressed image data having an arrangement following the respective compressed pixel positions corresponding to them, and some pixels of the same section overlap with the specified position of the predetermined section to be compressed. Obtained for each shift condition that has been shifted to a plurality of positions while satisfying the conditions, the compressed pixel data corresponding to the compressed image data under each shift condition are combined into one determination target compression data, and each determination target compression data A display screen inspection method characterized in that a defect in display characteristics is determined and detected for each compressed pixel by comparing with the reference data. 2. The method of inspecting a display screen according to claim 1, wherein in the synthesis, an average value is obtained from corresponding compressed pixel data .   2. The display screen inspection method according to claim 1, wherein the synthesis obtains one of a median value, a maximum value, and a minimum value from corresponding compressed pixel data.   Detection means for detecting predetermined display characteristics for each pixel constituting the display screen, storage means for storing each detected data so that it can be handled with vertical and horizontal addresses corresponding to the arrangement of each pixel, and stored in the storage means In addition, each set of pixel data corresponding to a plurality of pixels included in a predetermined section is compressed into one data, and the compressed data is arrayed in accordance with the corresponding compressed pixel position. The compressed image data is generated for each shift condition in which the specified position of the predetermined section to be compressed is shifted to a plurality of positions while satisfying the condition that the same pixel exists repeatedly. Means for combining the compressed pixel data corresponding to the compressed image data under each shift condition into one determination target compressed data; Inspection device for a display screen, characterized in that a defect judging means for detecting and determining a defect of display characteristics for each compressed pixel is compared with a reference data object compressed data. 5. The display screen inspection apparatus according to claim 4, wherein the compressed image synthesizing unit obtains an average value of corresponding compressed pixel data .   5. The display screen inspection apparatus according to claim 4, wherein the compressed image synthesizing unit obtains one of a median value, a maximum value, and a minimum value of corresponding compressed pixel data.   A method for inspecting a plasma display panel, comprising: detecting defects in the plasma display panel using the display screen inspection method according to claim 1.

Images