Image fusion method, system and device for product defect sample and storage medium

By generating high-quality and diverse product defect samples through image fusion technology, the problem of scarce defect samples in existing technologies is solved, and the generalization ability and robustness of deep learning models are improved, making them suitable for automatic detection of various industrial products.

CN122243772APending Publication Date: 2026-06-19SHANGHAI WESTWELL INFORMATION & TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
SHANGHAI WESTWELL INFORMATION & TECH CO LTD
Filing Date
2026-03-05
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

Existing technologies struggle to generate high-quality, diverse product defect samples, resulting in insufficient generalization ability of deep learning models in industrial defect detection. This is especially true under conditions of scarce and unbalanced real data, making it difficult to meet the demands for high-precision detection.

Method used

By acquiring a first product image containing defects and a second product image without defects, image recognition and segmentation are performed. After randomly selecting local images and performing color and texture correction, a gradient domain fusion algorithm is used to integrate the local images into the effective region of the second product image, generating a large number of high-quality defect samples.

Benefits of technology

The generated defect samples are realistic and diverse, significantly improving the model's generalization ability and robustness. The entire process is automated and requires no manual annotation, making it suitable for surface defect detection of various industrial products.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122243772A_ABST
    Figure CN122243772A_ABST
Patent Text Reader

Abstract

This invention provides an image fusion method, system, device, and storage medium for product defect samples. The method includes: acquiring a first product image containing defects as a first image set; segmenting and identifying various local regions of the defective product and locating the local region where the product defect is located; acquiring a second product image as a fusion background as a second image set; segmenting and identifying various local regions of the target product and an effective region for defining the allowed location of defects; randomly selecting a first product image and a second product image; extracting a local image of the product defect from the first product image; and performing color and texture correction on the local image based on the second product image; and performing image fusion to generate product defect samples. This invention can automatically generate a large number of high-quality, scene-adapted, and physically plausible damaged images, while achieving zero manual annotation costs, greatly improving data preparation efficiency.
Need to check novelty before this filing date? Find Prior Art