一种基于C波段SAR的海浪波高反演算法
By performing radiometric correction and filtering on C-band SAR data and combining it with a deep neural network model, the problems of complex data preprocessing and limited application scope in existing technologies have been solved, achieving efficient and wide-ranging wave height inversion.
CN120974892BActive Publication Date: 2026-07-17海南省航天技术创新中心 +1
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 海南省航天技术创新中心
- Filing Date
- 2025-07-23
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Existing SAR wave height inversion methods suffer from complex data preprocessing, resulting in low computational efficiency, narrow application scope, and limited application range due to input data requirements.
Method used
A wave height inversion algorithm based on C-band SAR was adopted. By performing radiometric correction and mean filtering on L1A level observation data in QPS mode, and inputting HH and VV polarization data, a deep residual neural network and a deep fully connected neural network model were constructed to invert the wave height.
Benefits of technology
It improves the efficiency and accuracy of ocean wave height retrieval, expands the application scope, and can be more widely used in the retrieval of SAR ocean wave height data.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN120974892B_ABST
Abstract
本发明公开了一种基于C波段SAR海浪波高反演算法，涉及海浪波高反演领域，包括：选取近岸海域卫星数据中的模式数据；选取ERA5风场数据和波高数据，经时间匹配后对模式数据进行处理，选取出SAR影像子图像，随后与ERA5风场数据和波高数据进行空间匹配；利用空间匹配后的SAR影像子图像、SAR数据辅助信息文件、ERA5风场数据和波高数据获取训练数据构建训练数据集；构建反演模型，将训练数据输入反演模型，最终得到SAR海浪波高。本发明采用了全极化数据开展海浪波高反演，精度高，应用门槛高，可应用范围小；同时综合考虑了反演精度和可应用性问题，更多的使用SAR信息用于海浪反演的同时又保证了反演模型的应用范围。
Need to check novelty before this filing date? Find Prior Art