A high-precision plate color restoration method and system based on color card correction and residual regression
By combining color card calibration and residual regression with global color correction and a lightweight machine learning model, the problem of insufficient color correction accuracy of board materials caused by material differences is solved, achieving high-precision and computationally efficient color restoration, which is suitable for industrial visual inspection and printing management.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Filing Date
- 2026-06-12
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies lack the accuracy for color correction of boards due to material differences. Traditional color card correction methods are limited by material differences, while deep learning methods are computationally complex and require large amounts of data, making them difficult to apply in practice.
A method based on color card correction and residual regression is adopted. By using global color correction and a lightweight machine learning model to compensate for material differences, and combining adaptive multinomial regression and residual prediction model, high-precision color reproduction is achieved.
It significantly improves the color reproduction accuracy of the board material, has high computational efficiency, is suitable for resource-constrained equipment, is highly adaptable, and is applicable to industrial online inspection and printed matter management.
Smart Images
Figure CN122415758A_ABST