# US 10,872,407 B2 · Method and apparatus for detecting defects, and computer readable storage medium

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 10,872,407 B2
- Document ID: US10872407B2
- Status: Granted
- Kind: B2
- Application number: 16/189,352
- Filed: 2018-11-13
- Published: 2019-07-25
- Granted: 2020-12-22
- Earliest claimed: 2018-01-22
- Assignees: Boe Technology Group Co., Ltd.; Chongqing Boe Optoelectronics Technology Co., Ltd.
- Applicants: Boe Technology Group Co., Ltd.; Chongqing Boe Optoelectronics Technology Co., Ltd.
- Inventors: Zhengdong Xi; GyuHyun Lee; Chaoqiang Liu; Jianhong Wang; Jinshun Wang
- Web page: https://patentgrep.com/patent/US10872407B2

## CPC

- G06T 7/001: {using an image reference approach}
- G06T 7/0004: {Industrial image inspection}
- G06T 7/0008: {checking presence/absence}
- G06T 7/60: Analysis of geometric attributes
- G06T 2207/20076: Probabilistic image processing
- G06T 2207/20224: Image subtraction
- G06T 2207/30108: Industrial image inspection
- G06T 2207/30121: CRT, LCD or plasma display
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

A method of detecting defects includes: obtaining a gray-scale image of a substrate to be detected; obtaining a gray-scale difference g between different regions in the gray-scale image; calculating an estimated value of a film thickness difference d between the different regions in the gray-scale image according to the gray-scale difference g; and determining whether the substrate to be detected is qualified according to the estimated value of the film thickness difference d.

## Claims

### Claim 1

A method of detecting defects, comprising:
obtaining a gray-scale image of a substrate to be detected;
obtaining a gray-scale difference g between different regions in the gray-scale image;
calculating an estimated value of a film thickness difference d between the different regions in the gray-scale image according to the gray-scale difference g; and
determining whether the substrate to be detected is qualified according to the estimated value of the film thickness difference d,
wherein obtaining the gray-scale difference g between different regions in the gray-scale image, comprises:
determining whether there is a gray-scale abnormal region in the gray-scale image, wherein the gray-scale abnormal region is a region in which a gray-scale value is too large or too small in the gray-scale image;
defining the gray-scale abnormal region and a gray-scale normal region in the gray-scale image in response to determining that there is the gray-scale abnormal region in the gray-scale image, and then proceeding to a next step;
determining that the substrate to be detected is qualified in response to determining that there is no gray-scale abnormal region in the gray-scale image; and
calculating the gray-scale difference g between the gray-scale abnormal region and the gray-scale normal region, and
wherein calculating the estimated value of the film thickness difference d between the different regions in the gray-scale image according to the gray-scale difference g, comprises:
calculating an estimated value of a film thickness difference d between the gray-scale abnormal region and the gray-scale normal region according to the gray-scale difference g.

### Claim 2 (depends on 1)

The method of detecting defects according to claim 1, wherein calculating the estimated value of the film thickness difference d between the gray-scale abnormal region and the gray-scale normal region according to the gray-scale difference g, comprises:
obtaining a relationship coefficient x representing a relationship between a sample film thickness difference and a sample gray-scale difference in sample data; and
calculating the estimated value of the film thickness difference d to be a product of the relationship coefficient x and the gray-scale difference g.

### Claim 3 (depends on 2)

The method of detecting defects according to claim 2, wherein obtaining the relationship coefficient x representing the relationship between the sample film thickness difference and the sample gray-scale difference in sample data, comprises:
collecting N sets of sample data, each set of sample data comprising a sample film thickness difference di and a sample gray-scale difference gi corresponding to a sample film thickness difference di, wherein N is greater than or equal to 2, i ranges from 1 to N, and d1˜dN differ from one another;
calculating a sample relationship coefficient xi of each set of sample data by dividing the sample film thickness difference di by the sample gray-scale difference gi, wherein the sample relationship coefficient xi represents a relationship between the sample film thickness difference di and the sample gray-scale difference gi in a corresponding set of sample data; and
calculating the relationship coefficient x according to sample relationship coefficients x1˜xN of the N sets of sample data.

### Claim 4 (depends on 3)

The method of detecting defects according to claim 3, wherein calculating the relationship coefficient x according to sample relationship coefficients x1˜xN of the N sets of sample data, comprises:
setting at least one determination range of the gray-scale difference, a collection of the at least one determination range being a continuous numerical range covering sample gray-scale differences g1˜gN of the N sets of sample data; and
calculating an average value of sample relationship coefficients corresponding to sample gray-scale differences falling into each of the at least one determination range to obtain a relationship coefficient x of a corresponding determination range.

### Claim 5 (depends on 4)

The method of detecting defects according to claim 4, wherein before calculating the estimated value of the film thickness difference d to be a product of the relationship coefficient x and the gray-scale difference g, the method further comprises:
determining a determination range into which the gray-scale difference g falls, and then determining a relationship coefficient x corresponding to the determination range.

### Claim 6 (depends on 1)

The method of detecting defects according to claim 1, wherein determining whether the substrate to be detected is qualified according to the estimated value of the film thickness difference d, comprises:
determining a fluctuation range, which ranges from (a difference between d and y) to (a sum of d and y), of the estimated value of the film thickness difference d, wherein y is a compensation coefficient representing an allowable error of the estimated value of the film thickness difference d relative to an actual value of the film thickness difference d′ of the substrate to be detected;
comparing the fluctuation range with a pre-set threshold value of the film thickness difference dmax, wherein the threshold value of the film thickness difference dmax is a maximum film thickness difference allowed for the substrate to be qualified;
determining that the substrate to be detected is qualified in response to the sum of d and y being smaller than or equal to dmax;
determining that the substrate to be detected is unqualified in response to the difference between d and y being greater than dmax;
measuring the actual value of the film thickness difference d′ in response to the difference between d and y being smaller than or equal to dmax and the sum of d and y being greater than dmax, and comparing the actual value of the film thickness difference d′ with the threshold value of the film thickness difference dmax;
determining that the substrate to be detected is qualified in response to d′ being smaller than or equal to dmax;
determining that the substrate to be detected is unqualified in response to d′ being greater than dmax.

### Claim 7 (depends on 1)

The method of detecting defects according to claim 1, wherein obtaining the gray-scale image of the substrate to be detected, comprises:
capturing an image of the substrate to be detected to obtain a captured image; and
converting the captured image into a gray-scale image.

### Claim 8 (depends on 1)

The method of detecting defects according to claim 1, wherein determining whether there is a gray-scale abnormal region in the gray-scale image, wherein the gray-scale abnormal region is a region in which a gray-scale value is too large or too small in the gray-scale image; defining the gray-scale abnormal region and a gray-scale normal region in the gray-scale image in response to determining that there is a gray-scale abnormal region in the gray-scale image, and then proceeding to a next step; determining if the substrate to be detected is qualified in response to determining that there is no gray-scale abnormal region in the gray-scale image, comprises:
obtaining a gray-scale value of each image point in the gray-scale image;
determining whether the gray-scale value of each image point is normal;
determining that there is a gray-scale abnormal region in the gray-scale image in response to determining that a gray-scale value of an image point is abnormal, and defining a region corresponding to image points with abnormal gray-scale values as a gray-scale abnormal region, and a region corresponding to image points with normal gray-scale values as a gray-scale normal region, and then proceeding to the step of calculating the gray-scale difference g between the gray-scale abnormal region and the gray-scale normal region; and
determining that there is no gray-scale abnormal region in the gray-scale image in response to determining that ray-scale values of all image points are normal, and then determining that the substrate to be detected is qualified.

### Claim 9 (depends on 8)

The method of detecting defects according to claim 8, wherein
each image point comprises a single pixel in the gray-scale image; or,
each image point comprises at least a column of pixels in the gray-scale image; or,
each image point comprises at least a row of pixels in the gray-scale image.

### Claim 10 (depends on 1)

The method of detecting defects according to claim 1, wherein calculating the gray-scale difference g between the gray-scale abnormal region and the gray-scale normal region, comprises:
determining a maximum gray-scale value in the gray-scale values of image points with abnormal gray-scale values in a case where gray-scale values of image points with abnormal gray-scale values are larger than gray-scale values of image points with normal gray-scale values, and calculating an average gray-scale value of image points with normal gray-scale values;
determining a minimum gray-scale value in the gray-scale values of image points with abnormal gray-scale values in a case where the gray-scale values of image points with abnormal gray-scale values are smaller than the gray-scale values of image points with normal gray-scale values, and calculating an average gray-scale value of image points with normal gray-scale values; and
subtracting the average gray-scale value from the maximum gray-scale value or subtracting the minimum gray-scale value from the average gray-scale value to obtain the gray-scale difference g.

### Claim 11 (depends on 1)

A non-transitory computer-readable storage medium storing executable instructions that, when executed by an apparatus for detecting defects, cause the apparatus to perform steps of the method of detecting defects according to claim 1.

### Claim 12

An apparatus for detecting defects, comprising:
a gray-scale image acquisition device configured to obtain a gray-scale image of a substrate to be detected;
a processor connected to the gray-scale image acquisition device, the processor being configured to obtain a gray-scale difference g between different regions in the gray-scale image, calculate an estimated value of a film thickness difference d between the different regions in the gray-scale image according to the gray-scale difference g, and determine whether the substrate to be detected is qualified according to the estimated value of the film thickness difference d,
wherein the processor is configured to:
determine whether there is a gray-scale abnormal region in the gray-scale image, wherein the gray-scale abnormal region is a region in which a gray-scale value is too large or too small in the gray-scale image;
define the gray-scale abnormal region and a gray-scale normal region in the gray-scale image in response to determining that there is the gray-scale abnormal region in the gray-scale image;
determine that the substrate to be detected is qualified in response to determining that there is no gray-scale abnormal regions in the gray-scale image;
calculate the gray-scale difference g between the gray-scale abnormal region and the gray-scale normal region;
calculate an estimated value of a film thickness difference d between the gray-scale abnormal region and the gray-scale normal region according to the gray-scale difference g; and
determine whether the substrate to be detected is qualified according to the estimated value of the film thickness difference d.

### Claim 13 (depends on 12)

The apparatus for detecting defects according to claim 12, further comprising a memory connected to the processor, the memory being configured to store a relationship coefficient x, wherein the relationship coefficient x represents a relationship between a sample film thickness difference and a sample gray-scale difference in sample data; and
the processor is configured to calculate a gray-scale difference g between the gray-scale abnormal region and the gray-scale normal region, read the relationship coefficient x from the memory, and calculate the estimated value of the film thickness difference d between the gray-scale abnormal region and the gray-scale normal region to be a product of the relationship coefficient x and the gray-scale difference g.

### Claim 14 (depends on 12)

The apparatus for detecting defects according to claim 12, wherein the processor is further configured to determine a defect type of a corresponding region of the substrate to be detected according to a distribution of image points in the gray-scale abnormal region.

### Claim 15 (depends on 12)

The apparatus for detecting defects according to claim 12, further comprising an alarm connected to the processor, the alarm being configured to send an alarm in response to determining that the substrate to be detected is unqualified.

### Claim 16 (depends on 12)

The apparatus for detecting defects according to claim 12, wherein the gray-scale image acquisition device comprises a linear camera.