基于参数化稀疏重建的舰船目标微多普勒特征提取方法研究

项目名称： 基于参数化稀疏重建的舰船目标微多普勒特征提取方法研究

项目编号： No.41271011

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 天文学、地球科学

项目作者： 李刚

作者单位： 清华大学

项目金额： 71万元

中文摘要： 舰船分类与识别是微波遥感领域的重要任务之一，直接服务于海面监视与预警这一国家重大应用需求。提高舰船分类与识别的正确率关键在于提取舰船的标志性特征，如雷达天线转动速度、发动机振动频率等，提取这些特征的最有效手段是微多普勒分析。由于舰船的整体运动、局部转动与振动都是非合作的，现有的微多普勒分析方法很难实现舰船目标微多普勒特征的快速准确提取。为了解决这一难题，在青年科学基金项目研究成果“参数化稀疏表征理论”和“成簇稀疏约束模型”的基础上，本项目将研究基于参数化稀疏重建的舰船目标微多普勒特征提取新方法。研究内容包括：舰船目标微多普勒产生机理与信号模型、舰船目标微多普勒信号的参数化稀疏表征方法、舰船目标微多普勒特征的参数化稀疏重建算法。通过本项目的研究，实现低数据量、高精度的舰船目标微多普勒特征提取，在微波遥感领域的关键科学问题上取得原创性成果，为海面监视与预警这一国家重大需求提供理论支持。

中文关键词： 微多普勒雷达；舰船目标识别；稀疏信号处理；；

英文摘要： Ship classification and identification is one of important tasks in the field of microwave remote sensing, which provides key information for sea surface monitoring and early-warning. Feature extraction is the key to improve the success probability of ship classification and identification, and Micro-Doppler analysis is effective to extract some features such as radar antenna rotation speed and engine vibration frequency. Because the rigid motion of ships and the rotation/vibration of local parts are non-cooperative, existing methods of Micro-Doppler analysis are difficult to fast and accurately extract the Micro-Doppler feature of ships. In order to solve this problem, by using the research results of Young Scientist Funding, "parametric sparse representation theory" and "clustering sparse model", this project will study new methods of Micro-Doppler feature extraction of ships based on parametric sparse recovery. The study includes: the principle of Micro-Doppler effect of ships and the signal model, the parametric sparse representation of Micro-Doppler signals of ships, and the parametric sparse recovery of Micro-Doppler signals of ships. This project will implement the Micro-Doppler feature extraction with high precision from less measurements and present originality achievements in the field of microwave rem

英文关键词： micro-Doppler radar；ship recognition；sparse signal processing；；