基于多尺度Transformer的海洋遥感图像大气校正方法
By constructing a multi-scale Transformer model and a composite loss function with physical soft constraints, the computational bottleneck and insufficient generalization ability of atmospheric correction in large-scale marine environments in existing technologies are solved, achieving efficient and accurate atmospheric correction, which is suitable for global large-scale satellite remote sensing data processing.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- QILU UNIVERSITY OF TECHNOLOGY (SHANDONG ACADEMY OF SCIENCES)
- Filing Date
- 2026-05-09
- Publication Date
- 2026-07-17
AI Technical Summary
Existing atmospheric correction methods suffer from computational bottlenecks and insufficient generalization capabilities in large-scale, automated operational applications, especially in complex marine environments where high-precision and robust atmospheric correction is difficult to achieve.
An atmospheric correction method for marine remote sensing images based on multi-scale Transformer is adopted. By constructing a multi-scale Transformer model with parallel convolutional branches and combining it with a composite loss function with physical soft constraints, end-to-end atmospheric correction inference is performed. The model is trained using multi-source remote sensing observation data and auxiliary parameters to achieve high-precision and robust atmospheric correction.
It achieves high-precision, robust, and physically consistent atmospheric correction in complex, large-scale marine environments, significantly improving computational efficiency, taking into account both global features and local high-frequency details, possessing cross-temporal and spatial generalization capabilities, and ensuring the physical consistency and numerical fidelity of the output results.
Smart Images
Figure CN122175812B_ABST