一种遥感图像语义分割优化方法
By optimizing remote sensing image segmentation using confidence gating and undirected graphical models, the problems of computational redundancy and manual dependence are solved, achieving efficient and automated semantic segmentation of remote sensing images and improving the computational efficiency and accuracy of image segmentation.
CN120913208BActive Publication Date: 2026-07-17AEROSPACE DONGFANGHONG SATELLITE
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- AEROSPACE DONGFANGHONG SATELLITE
- Filing Date
- 2025-07-31
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Existing methods for optimizing semantic segmentation of remote sensing images suffer from computational redundancy, strong reliance on manual intervention, and insufficient optimization of local details.
Method used
By using a confidence gating mechanism and an undirected graph model, an undirected graph with image pixels as nodes and pixel adjacencies as edges is constructed. Logits propagation and iterative updates are then performed to optimize image segmentation labels.
Benefits of technology
It achieves low-overhead, automated, and high-precision semantic segmentation of remote sensing images, improving computational efficiency and detail accuracy, and is suitable for real-time high-resolution remote sensing image processing.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN120913208B_ABST
Abstract
本发明公开了一种遥感图像语义分割优化方法，包括：对遥感图像进行语义分割，得到分类逻辑值图像；提取得到分类逻辑值，将分类逻辑值转换为类别概率，并进行全图置信度量化；将分类逻辑值图像划分为保留区域和待优化区域；构建以图像像素为节点、像素邻接为边的无向图模型；针对待优化区域节点开展logits传播，并迭代更新节点logits分布；根据更新后的节点logits分布，输出优化后的图像分割标签，完成遥感图像的识别处理。本发明所述方法，解决了现有遥感图像语义分割优化方法普遍存在计算冗余、人工依赖性强、局部细节优化不足等问题，具有低开销、自动化、高精度等优点。
Need to check novelty before this filing date? Find Prior Art