基于视觉先验学习和混合因子分析的极化SAR图像识别与分类

本项目针对当前极化SAR数据处理只关注极化散射特性，而没有考虑人类视觉感知特性的问题，利用视觉先验，将压缩感知理论和视觉注意理论相结合，提出高分辨极化SAR图像的稀疏模型与自适应学习字典构造方法,建立极化SAR图像典型目标的压缩感知域多尺度统计模型和目标观测特征一体化提取方法；将混合因子分析和所提出的稀疏模型相结合，建立极化SAR图像目标识别与分类，和基于压缩谱聚类的SAR图像地物分类和聚类，并用RADARSAT 2和PALSAR的全极化SAR数据验证其有效性。期望在提高极化SAR图像目标识别与分类效果的同时，能进一步完善和促进压缩感知理论的研究和应用。研究成果在本领域重要期刊和会议上发表论文15-20篇，申报国家发明专利6-8项，联合培养博士、硕士5-8名。

This project aims at the problem of current polarimetric SAR data processing not considering visual perception characteristic but only using polarimetric scattering, firstly, high resolution polarimetric SAR sparse model and adaptive learning dictionary construction will be proposed combined with compressed sensing and visual attention theory using visual prior, and the multiscale statistical model of polarimetric SAR image object in compressed sensing domain and the integrative observed feature extraction will be also proposed. Secondly, Polarimetric SAR image object recognition and classification and SAR image terrain classification and clustering combined with Mixture of Factor Analyzers and the sparse model will be proposed. We validate the effectiveness of the proposed method by the full-polarimetric data of RADARSAT 2 and PALSAR. We hope our methods not only to improve performance of Polarimetric SAR image object recognition and classification, but also further boosting and carving out compressed sensing theory and its applications. We will publish 10-15 journals and conferences, apply 6-8 patents and bring up 5-8 Ph.Ds and masters.