OLED display screen microdefect automatic detection and classification method and system

By fusing local frequency domain features with global contextual attention and employing a meta-learning cross-screen adaptation framework, this approach addresses the issues of insufficient feature extraction and weak adaptability in OLED display micro-defect detection. It achieves high-precision, low-cost automatic detection and classification, enabling rapid adaptation to different screen models and high-confidence decision-making.

CN122134622APending Publication Date: 2026-06-02SHENZHEN KELAI INTELLIGENT DISPLAY CO LTD
0 Cites 2 Cited by

Patent Information

Application Number
CN202610082077.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-01-21
Publication Date
2026-06-02

AI Technical Summary

Technical Problem

Existing micro-defect detection technologies for OLED displays have shortcomings in terms of insufficient feature extraction, weak adaptability to different screens, and difficulty in assessing the reliability of detection results. In particular, they are not effective in identifying textured and periodic defects. Furthermore, when adapting the model to new screen models, a large amount of data needs to be re-annotated, which is costly and requires a large amount of manual re-inspection.

Method used

By employing a fusion mechanism of local frequency domain features and global contextual attention, combined with a meta-learning cross-screen adaptation framework and a dual-path uncertainty quantization strategy, and through pixel response modeling, differential processing, multi-scale residual feature enhancement, frequency domain filter enhancement, and meta-learning mechanism, we can achieve rapid adaptation and high-confidence decision-making for different screen models.

Benefits of technology

It significantly improves the ability to characterize micro-defects and classify them, reduces labeling costs and the burden of manual review, and forms a high-precision, highly generalizable, and self-evolving OLED micro-defect detection system.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses an OLED display screen microdefect automatic detection and classification method and system, and relates to the technical field of display panel quality detection.The OLED display screen microdefect automatic detection and classification method comprises the steps that an OLED display image is collected and preprocessed, pixel response modeling is carried out based on the preprocessed image, and a pixel response model is obtained; an ideal background image is generated and differential processing and multi-scale residual feature enhancement are carried out on the ideal background image to obtain an enhanced residual image; segmenting the enhanced residual image to extract a candidate defect region set, extracting global features, and based on the candidate defect region set, performing frequency domain enhancement on the global features through a local frequency domain prior guide complex filter to obtain enhanced global context features; and on the basis of the enhanced global context features, new screen type adaptive parameters are obtained by using a meta-learning mechanism to carry out defect category probability prediction, and uncertainty estimation is carried out on the basis of a defect category probability prediction result. According to the OLED display screen defect detection and classification method, OLED display screen defect detection and classification which are more accurate, more flexible and capable of evaluating credibility can be realized.
Need to check novelty before this filing date? Find Prior Art