一种面向多源图像的自动化采集与质量增强方法及系统
By using RANSAC registration and wavelet decomposition fusion technology driven by synchronous acquisition and quality scoring, the problem of limited image quality improvement in automated processing of multi-source images is solved, and high-precision image registration and quality enhancement effects are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUNAN INST OF INFORMATION TECH
- Filing Date
- 2026-04-28
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies fail to effectively utilize complementary information between multiple images in automated processing of multi-source images, resulting in limited improvement in image quality, especially in terms of image registration accuracy and the lack of effective decision-making mechanisms for improving global contours and local details.
By simultaneously acquiring multi-source images, extracting local spatial features and global statistical features, calculating quality scores, performing weighted registration using the RANSAC algorithm, and generating enhanced images through multi-layer wavelet decomposition and fusion of high-frequency and low-frequency coefficients driven by quality scores.
It achieves high-precision registration and quality enhancement of multi-source images, improving image contrast, clarity and detail richness, and realizing a fully automated closed loop from acquisition to enhancement.
Smart Images
Figure CN122134774B_ABST