地物变化检测方法、装置及非易失性存储介质
By acquiring and segmenting ground feature images from remote sensing images and utilizing feature difference detection methods, the problem of low accuracy in detecting ground feature changes in remote sensing images is solved, and high-precision ground feature change detection is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- STATE GRID BEIJING ELECTRIC POWER CO
- Filing Date
- 2022-06-30
- Publication Date
- 2026-07-17
AI Technical Summary
Existing methods for detecting changes in land features have low accuracy when detecting changes in land features in remote sensing images, and their reliance on human experience leads to strong subjectivity, making it impossible to effectively improve detection accuracy.
By acquiring first and second remote sensing images, segmentation is performed according to the target segmentation scale to determine the feature differences of the ground object images. A segmentation quality evaluation function is constructed using the gray-level co-occurrence matrix texture information entropy and the absolute value of the difference between the mean and mean of the object spectrum. Combining fuzzy similarity and angular texture features, feature fusion is performed using the image two-dimensional entropy separation threshold algorithm and the PCNN algorithm to achieve ground object change detection.
It improves the accuracy of ground feature change detection in the field of remote sensing, enabling accurate detection of whether ground features have changed, reducing subjectivity and improving detection efficiency.
Smart Images
Figure CN115205700B_ABST