高距离分辨雷达目标特征提取的参数化方法

【作者】 陈黎霞；

【导师】 裴炳南；

【作者基本信息】 郑州大学 ， 通信与信息系统， 2005， 硕士

【摘要】 本文围绕国家自然科学基金项目“基于几何散射理论提取高距离分辨雷达目标特征方法研究”(项目编号:60272059),针对高距离分辨雷达像的特征提取的参数化方法及其相关问题作了一些粗浅的研究。 论文首先简单介绍了雷达目标识别技术的发展状况,特征提取是雷达目标识别的关键环节。然后在下面两章研究了Relax算法和匹配追踪算法。Relax算法具有提取参数精度高,数值可靠,对散射点的选择具有一定的鲁棒性等优点,但是此算法是基于简单散射点模型的,没有利用HRRP中含有散射点的几何形状这一特征信息。匹配追踪算法比较简便逼近效果较好,但是超完备词典的选择比较困难,且它使用贪心算法,在每一步迭代都力图尽量捕捉信号的能量,因此算法的主要缺点是计算量过于庞大。 为充分利用HRRP中散射中心的几何特征信息,并避免匹配追踪算法中基函数选择的困难,基于几何散射理论,本文采用GTD模型,在数值计算上借鉴Relax算法思想,给出了提取目标散射中心位置、强度以及几何形状参数信息的RIP算法思想,尝试开辟一个特征提取和目标识别的新思路。并结合ISAR外场实测数据及计算机模拟高距离分辨雷达回波数据进行了大量的仿真试验。更多还原

【Abstract】 Some specific techniques for ATR by a high range resolution radar are researched in feature extraction, which are parts of a project—"Search into Method for HRR Radar Target Feature Extraction Based on Geometrical Theory of Diffraction" supported by National Science Foundation of China.At the beginning, the thesis makes a brief introduction to the development of radar target recognition technique, and feature extraction is the pivotal point. Then the follow two chapters study Relax algorithm and Matching Pursuit algorithm. The parameters extracted by Relax algorithm are correct and dependable, but the geometrical information included in the high range resolution profile has not been utilized by Relax algorithm because Relax algorithm is based on simple scattering-centers model. Matching Pursuit algorithm is simple, however, the choice of over-complete dictionary is very difficult, and it has enormous calculation due to its greedy iterative algorithm.So, in order to fully utilize the geometrical information included in the scattering points, and to avoid the difficulty of the choice of basis in Matching Pursuit algorithm, the thesis put forward a new method of feature extraction based on Geometrical Theory of Diffraction model to extract the geometrical parameter of the scattering centers except the intensity and position parameters. In the end, simulation results of the proposed algorithm are given.更多还原