一种文档图像多缺陷自适应修复与增强的方法及系统
By employing a step-by-step multi-defect perception and dynamic priority scheduling mechanism, combined with three-level progressive geometric correction and pixel-level defect repair, the problem of multiple complex defects in document images is solved, achieving efficient and accurate document image repair and enhancement, and improving text clarity and image purity.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SICHUAN JISU POWER TECH CO LTD
- Filing Date
- 2026-06-22
- Publication Date
- 2026-07-17
AI Technical Summary
Existing document image restoration and enhancement technologies cannot handle multiple complex defects simultaneously, resulting in problems such as text edge blurring, stroke blurring, text deformation, table distortion, abnormal exposure, and color deviation, making it impossible to achieve pixel-level detail restoration and efficient digital processing.
By employing a step-by-step multi-defect perception and dynamic priority scheduling mechanism, and through three-level progressive geometric correction and pixel-level defect repair, combined with super-resolution reconstruction and closed-loop evaluation, we can achieve adaptive repair and enhancement of multiple defects in document images.
It achieves collaborative repair of complex defects such as 3D surface wrinkles, geometric distortion, multiple types of blur, low resolution, exposure anomalies, and color deviation, improving text clarity and recognizability, ensuring image purity and computational efficiency, and adapting to the digital needs of various document scenarios.
Smart Images
Figure CN122415397A_ABST