基于分层稀疏表示的微动目标ISAR三维层析成像技术

项目名称： 基于分层稀疏表示的微动目标ISAR三维层析成像技术

项目编号： No.61501474

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 无线电电子学、电信技术

项目作者： 金光虎

作者单位： 中国人民解放军国防科技大学

项目金额： 19万元

中文摘要： 起伏、自旋、翻滚、进动等微动使得目标ISAR成像时空间谱呈现三维曲面特性，二维/三维成像难度大、效果差。本项目利用微动的三维转动特点，提出一种空间支撑谱三维曲面的分层节省参数表示方法,将3D-ISAR成像转化为ISAR切片成像和三维ISAR层析成像，把复杂的稀疏字典构建转化为两个一维成像稀疏字典构建，基于稀疏表示理论，实现ISAR切片成像和高度维层析成像。项目变微动从传统观念的成像不利因素为三维成像的有利因素，提出了分层、分步ISAR三维层析成像思路和基于稀疏表示理论的ISAR三维层析成像方法，相比于直接三维成像，该方法计算量小、稳健性高，可望在成像效果、成像效率及稳健性上获得突破。

中文关键词： ISAR成像；三维成像；微动；空间谱分层；稀疏表示

英文摘要： Micro-motion such as waving，spinning，rolling，precessing makes the space spectra appear as three-dimensional curved surface. It is hard to get clear 2D/3D ISAR image. Base on the periodicity of micro-motion, we proposed to delaminate space supporting spectra into several layers. Therefore, the 3D ISAR imaging is translated into Slice-ISAR imaging and 3D tomography ISAR imaging. And the uniformed sparse reprensentation is translated into two sparse reprensentations for single dimension imaging. 3D ISAR tomography imaging is realized based on sparse representation theory lastly. The project turned the micro-motion from disadvantage in trandistional ISAR imaging to advantage for 3D-ISAR imaging. 3D ISAR tomography imaging based on sparse representation theory by two steps process is proposed. Compared to direct 3D imagine, the algorithm has the merits of high efficiency, good robustness on which the project is likely to get breakthrough.

英文关键词： ISAR imaging;3D imaging;micro-motion;spectra delaminating;sparse representation