JPH04336384A - Pattern inspection device - Google Patents

Abstract

PURPOSE:To enable automatic inspection by performing a gradation expanding and reducing process as a preprocess for a binarization processing so as to remove the influence of disturbance. CONSTITUTION:After a gradation image is inputted, the preprocess by the gradation expanding and reducing process is performed and this gradation expanding and reducing process is carried out in proper combination according to the density distribution state of a pattern part to be inspected and a background. When the brightness value distribution of the pattern part to be inspected is lower than that of the background, the inspection pattern is erased from the image by the reducing process from the expanding process to generate a gradation image consisting of only the background. When the light and dark distribution state is high, on the other hand, the expanding process is performed from the reducing process. The image and original image are used to perform arithmetic between the images. At this time, an illumination irregularity and a color irregularity in the image are removed. Therefore, the gradation image after the preprocess is subjected to binarization processing by using a single threshold value. Further, the area, etc., of the pattern part to be inspected is analyzed to perform the inspection.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a display pattern inspection device for liquid crystal panels and a circuit pattern inspection device for substrates.

0002

[Prior Art] A conventional pattern inspection device performs a binarization process on a grayscale image captured in an image memory using a single or locally set threshold value, and then performs a binarization process to distinguish between background and inspection images. The target pattern was isolated and inspected.

0003

[Problems to be Solved by the Invention] However, in conventional pattern inspection devices, when there is uneven brightness due to illumination, uneven color in the pattern to be inspected, etc. in the captured image, the brightness distribution state becomes unstable. In the case of a change, there is a problem that binarization processing using a preset threshold value cannot separate only the pattern portion to be inspected, and defects cannot be inspected due to area or the like.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to solve the above-mentioned problems of the conventional art, and to make it possible to distinguish between the pattern part to be inspected and the pattern part to be inspected without being affected by the unevenness of brightness caused by illumination or the uneven color of the pattern part to be inspected. It is an object of the present invention to provide a pattern inspection device that can create a binary image separated from the background, analyze the image, and perform defect inspection based on area and the like.

[0005]

[Means for Solving the Problems] In order to solve the above problems, the present invention has a configuration in which gray scale expansion and reduction processing is performed as pre-processing for binarization processing in order to remove the influence of disturbance, and the background and inspection Stable separation of target pattern parts is now possible.

[0006]

[Operation] In the pattern inspection device configured as described above, the gradation expansion/contraction processing works, and even images with uneven illumination and coloring can be binarized without being affected by these disturbances. It becomes possible to separate the background and the pattern portion to be inspected, and pattern inspection based on area, etc. becomes possible.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Examples of the present invention will be described below with reference to the drawings. FIG. 1 is an explanatory diagram showing an inspection procedure of a pattern inspection apparatus according to the present invention. As shown in FIG. 1, in the present invention, after capturing a grayscale image, preprocessing is performed using grayscale expansion/reduction processing, and inter-image calculation is performed from the grayscale image before processing and the grayscale image after processing. Create a grayscale image with only the pattern portion extracted. The image is subjected to binarization processing to separate the background and the pattern part to be inspected, and the area etc. of the pattern part to be inspected is analyzed.
Conduct an inspection.

FIG. 2 is an explanatory diagram showing an embodiment of a pattern inspection apparatus according to the present invention. In the same figure, when a sample 2 whose pattern is to be inspected sent on a transport rail 1 passes in front of a photosensor 3, a control signal is sent from the photosensor 3 to an electrical controller 4, and the sample 2 is inspected by a TV camera 5. It stops directly below. At this time, the TV camera 5 above the sample 2 captures an image of the sample 2, and the image processing unit 6 inputs the image. The image processing section 6 performs gradation expansion/reduction processing on the captured image, inspects the pattern of the sample 2, and transfers the results to the electronic controller 4. If the test result is good, the sample 2 is sent to the next step by the transport rail 1, and if the test result is bad, the sample 2 is sent to the electrical controller 4.
The sample 2 is carried out by the carry-out machine 7 to the carry-out table 8 according to instructions from the sample 2, and is distinguished as a defective product.

Next, the grayscale expansion/reduction process used before creating a binary image in the present invention will be explained. FIG. 3 is an explanatory diagram of the shading expansion process. The gradation expansion process refers to a process in which the brightness value of a pixel within the designated shape is replaced by the highest brightness value within the range of the designated shape 10 centered on each pixel. Grayscale image 9 is an image before expansion processing. Each square corresponds to one pixel, and the numbers represent the brightness value of each pixel. A designated shape 10 is sequentially applied to each pixel constituting this grayscale image 9 by raster scanning, and the brightness values of all pixels within the designated shape 10 are replaced with the value with the highest brightness value among them. The image after gradation expansion processing in which this operation is performed for all pixels on the screen is the gradation image 11.

FIG. 4 is an explanatory diagram of the shading reduction process. Shade reduction processing is a specified shape 13 centered on each pixel.
This refers to the process of replacing the brightness value of a pixel within a specified shape with the lowest brightness value within the range of . The grayscale image 12 is an image before reduction processing. Each square corresponds to one pixel, and the numbers represent the brightness value of each pixel. The specified shape 13 is centered around each pixel composing this grayscale image 12.
are sequentially applied by raster scanning, and the luminance value of the pixel within the specified shape 13 is replaced with the lowest luminance value among them. The image after the gradation reduction process in which this operation is performed for all pixels on the screen is the gradation image 14.

[0011] This shading expansion/reduction processing is executed in appropriate combinations depending on the density distribution state of the pattern portion to be inspected and the background. If the luminance value distribution of the pattern to be inspected is lower than the background, expansion processing and reduction processing are performed to erase the inspection pattern from the image and create a grayscale image of only the background. If the grayscale distribution state is reversed, the reduction process is followed by the expansion process. An inter-image operation (subtraction) is performed using this image and the original image. At this time, illumination unevenness and color unevenness included in the image are removed. Therefore, the grayscale image that has undergone this preprocessing can be binarized using a single threshold value.

0012

Effects of the Invention As explained above, even if the input image has uneven illumination or color unevenness in the pattern, this invention can remove them by the gradation expansion/reduction processing, thereby reducing the disturbance. This makes it possible to inspect samples that are difficult to binarize due to color unevenness. As a result, processes that conventionally relied on human visual inspection can be replaced with automatic inspection by machines, making it possible to improve accuracy, reduce labor costs, and improve productivity.

[Brief explanation of drawings]

FIG. 1 is an explanatory diagram showing an inspection procedure of an inspection apparatus according to the present invention.

FIG. 2 is an explanatory diagram showing an embodiment of the present invention.

FIG. 3 is an explanatory diagram showing a gradation expansion processing method used in the present invention.

FIG. 4 is an explanatory diagram showing a shading reduction processing method used in the present invention.

[Explanation of symbols]

1 Transport rail 2 Sample 3 Photo sensor 4 Electrical controller 5 TV camera 6 Image processing section 7 Unloading machine 8 Unloading table 9. Grayscale screen before expansion processing 10 Expansion processing specified shape 11. Grayscale screen after expansion processing 12. Grayscale screen before reduction processing 13　Reduction processing specified shape 14. Grayscale screen after reduction processing

Claims (1)

[Claims] 1. A pattern inspection device that separates a background from a pattern by binarizing a grayscale input image stored in an image memory in the device, and inspects the pattern for defects based on its area, etc. It is characterized by having an image processing unit that performs a grayscale expansion/reduction process as a preprocessing of the binarization process, and performs a pattern inspection using a binary image obtained by the binarization process after the grayscale expansion/reduction process. Pattern inspection equipment.