Fusion detail enhanced visual detection image defogging processing method and system
By analyzing visual feature partitioning and fog occlusion indices, and combining partition detection detail coefficients for adaptive dehazing, this method solves the problem of difficulty in coordinating detail enhancement and dehazing intensity in image dehazing methods, achieving accurate and uniform dehazing effects and reliable target detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHENZHEN YUEFEIYANG TECH CO LTD
- Filing Date
- 2026-03-18
- Publication Date
- 2026-06-12
AI Technical Summary
Existing image dehazing methods struggle to coordinate detail enhancement with dehazing intensity, leading to issues such as abrupt boundary changes and mismatch between fog area processing and detection targets in partitioned dehazing.
Visual feature partitioning is performed based on the detection target, partition detection features and partition detection detail coefficients are obtained, fog occlusion partitions are established through fog occlusion index analysis, and partition detection detail coefficients are used for mapping to identify the defogging sensitivity constraints of each partition and perform adaptive defogging processing.
It achieves accurate, uniform, and boundary-free dehazing, improving the visual quality of hazy images and the reliability of target detection.
Smart Images

Figure CN122199985A_ABST