CN-121982378-A - Display screen defect detection method, device and system, electronic equipment and storage medium

CN121982378ACN 121982378 ACN121982378 ACN 121982378ACN-121982378-A

Abstract

The application discloses a display screen defect detection method, device and system, electronic equipment and a storage medium, and belongs to the technical field of defect detection. The method comprises the steps of obtaining a first image and a second image of a display screen, wherein the first image is an image of the display screen in a lighting state, the second image is an image of the display screen in a blanking state and irradiated by a light source, conducting image fusion on the first image and the second image to obtain a target fusion image, inputting the target fusion image into a target classification model to obtain a defect classification result output by the target classification model, wherein the defect classification result is used for representing whether defects of the target fusion image are appearance defects or not, the target classification model is a pre-trained deep learning classification network, and determining a defect detection result of the display screen based on image data of which the defect classification result is non-appearance defects. The method can accurately remove external interference factors such as dust, remarkably improve the detection through rate of the product and effectively reduce the over-killing rate.

Inventors

XING ZHIWEI
SHI GUANGJUN
ZHAO YAN

Assignees

苏州凌云光工业智能技术有限公司
凌云光技术股份有限公司

Dates

Publication Date: 20260505
Application Date: 20251230

Claims (10)

1. A display screen defect detection method, comprising: Acquiring a first image and a second image of a display screen, wherein the first image is an image of the display screen in a lighting state, and the second image is an image of the display screen in a extinguishing state and irradiated by a light source; Performing image fusion on the first image and the second image to obtain a target fusion image; Inputting the target fusion image into a target classification model to obtain a defect classification result output by the target classification model, wherein the defect classification result is used for representing whether the defect of the target fusion image is an appearance defect or not, and the target classification model is a pre-trained deep learning classification network; And determining a defect detection result of the display screen based on the image data of which the defect classification result is a non-appearance defect.
2. The method for detecting a defect of a display screen according to claim 1, wherein the target fusion image includes at least one target fusion sub-image, and the performing image fusion on the first image and the second image to obtain the target fusion image includes: Performing defect extraction on the first image to obtain defect position information of the display screen; capturing the first image based on the defect position information to obtain a first defect subgraph, and capturing the second image based on the defect position information to obtain a second defect subgraph; And fusing the first defect subgraph and the second defect subgraph at the same defect position to obtain the target fused subgraph, wherein the defect classification result is used for representing whether the defect corresponding to the target fused subgraph is an appearance defect or not.
3. The display screen defect detection method of claim 2, wherein the first defect subgraph and the second defect subgraph are single-channel images and the target fusion subgraph is a three-channel image.
4. The display screen defect detection method of claim 1, wherein the appearance defects include dust, oil stains, and scratches.
5. The method of any one of claims 1-4, wherein the object classification model comprises a plurality of sub-modules connected in sequence, the sub-modules comprising a plurality of reverse residual blocks connected in sequence.
6. The display screen defect detection method of claim 5, wherein the reverse residual block comprises an inverted bottleneck structure and an attention mechanism module.
7. A display screen defect detection apparatus, comprising: the device comprises an acquisition module, a display screen display module and a display module, wherein the acquisition module is used for acquiring a first image and a second image of the display screen, the first image is an image of the display screen in a lighting state, and the second image is an image of the display screen in a extinction state and irradiated by a light source; The first processing module is used for carrying out image fusion on the first image and the second image to obtain a target fusion image; the second processing module is used for inputting the target fusion image into a target classification model to obtain a defect classification result output by the target classification model, wherein the defect classification result is used for representing whether the defect of the target fusion image is an appearance defect or not, and the target classification model is a pre-trained deep learning classification network; And the third processing module is used for determining the defect detection result of the display screen based on the image data of which the defect classification result is a non-appearance defect.
8. A display screen defect detection system, comprising: The device comprises a light source and an image acquisition device, wherein the image acquisition device is used for acquiring a first image and a second image of a display screen, the first image is an image of the display screen in a lighting state, and the second image is an image of the display screen in a extinguishing state and under the irradiation of the light source; And the data processing device is connected with the image acquisition equipment and is used for executing the display screen defect detection method according to any one of claims 1-6 and outputting a defect detection result of the display screen.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the display defect detection method of any of claims 1-6 when the program is executed by the processor.
10. A non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor implements the display screen defect detection method according to any of claims 1-6.

Description

Display screen defect detection method, device and system, electronic equipment and storage medium Technical Field The application belongs to the technical field of defect detection, and particularly relates to a display screen defect detection method, device and system, electronic equipment and a storage medium. Background In the lighting detection process of the display screen, the conventional detection method is to use an industrial camera system to shoot the display state of the display screen when the display screen is lighted and convert the display state into images, then use a vision processing system to process the images, detect various defects such as bright spots, dark spots, uneven display and the like, automatically judge according to preset defect standards, and obtain the conclusion that the display screen is good or defective. However, in the display screen lighting detection process, dust existing in the air or dirt on the surface of the display screen can be imaged on the camera, black or gray pixels are displayed and are easily detected as defects by the vision processing system, so that good products are misjudged as defective products, and the detection straight-through rate is reduced. Disclosure of Invention The present invention aims to solve at least one of the technical problems existing in the prior art. Therefore, the display screen defect detection method, device, system, electronic equipment and storage medium provided by the invention can accurately remove external interference factors such as dust, and the like, the defect detection result of the display screen is more accurate, the product detection through rate is improved, and the over-killing rate is effectively reduced. In a first aspect, the present application provides a method for detecting a display defect, where the method includes: Acquiring a first image and a second image of a display screen, wherein the first image is an image of the display screen in a lighting state, and the second image is an image of the display screen in a extinguishing state and irradiated by a light source; Performing image fusion on the first image and the second image to obtain a target fusion image; Inputting the target fusion image into a target classification model to obtain a defect classification result output by the target classification model, wherein the defect classification result is used for representing whether the defect of the target fusion image is an appearance defect or not, and the target classification model is a pre-trained deep learning classification network; And determining a defect detection result of the display screen based on the image data of which the defect classification result is a non-appearance defect. According to the display screen defect detection method, the first image of the display screen in the on state and the second image of the display screen in the off state and under the irradiation of the light source are fused to obtain the target fusion image, the target fusion image is processed by adopting the target classification model, whether defects on the display screen are appearance defects such as dust, greasy dirt and scratches or not is accurately judged by adopting the deep learning classification network, finally, the image data of non-appearance defects is adopted to obtain the defect detection result of the display screen, external interference factors such as dust can be accurately eliminated, the defect detection result of the display screen is more accurate, the product detection through rate is remarkably improved, and the over-killing rate is effectively reduced. According to one embodiment of the present application, the target fusion image includes at least one target fusion sub-image, and the image fusion is performed on the first image and the second image to obtain a target fusion image, including: Performing defect extraction on the first image to obtain defect position information of the display screen; capturing the first image based on the defect position information to obtain a first defect subgraph, and capturing the second image based on the defect position information to obtain a second defect subgraph; And fusing the first defect subgraph and the second defect subgraph at the same defect position to obtain the target fused subgraph, wherein the defect classification result is used for representing whether the defect corresponding to the target fused subgraph is an appearance defect or not. According to an embodiment of the present application, the first defect sub-graph and the second defect sub-graph are single-channel images, and the target fusion sub-graph is a three-channel image. According to one embodiment of the application, the appearance defects include dust, oil stains and scratches. According to one embodiment of the application, the object classification model comprises a plurality of sub-modules connected in sequence, the sub-modules comprising a plurali