摘要
高分辨距离像(HRRP)目标识别算法很多,在其利用高分辨距离像蕴含的目标结构信息的同时,也需要面对数据量巨大的难题。事实上,尽管高分辨距离像数据量巨大,但却是稀疏的,然而利用其稀疏特性进行识别的方法却不多。为此,提出了一种基于压缩感知稀疏表示方法实现目标识别的算法。该算法首先采用遗传正交匹配追踪(OMP)算法对一维距离像训练样本进行稀疏分解以获得类别字典,然后根据类别字典分析测试样本的重构误差实现目标识别。仿真实验证明,所提算法简捷、识别率更高,相较于常规算法识别率提高最多可达20%,并且在受到噪声干扰情况下依然能够稳健地识别目标。
There are many radar target recognition algorithms for high resolution range profile( HRRP). All of them use the target structure information embedded in HRRPs. However,it is difficult to extract and analyze such vast amounts of data. In fact,HRRPs are sparse,but less of radar target recognition algorithms employ the sparseness of HRRPs. Thus,a fast sparse representation algorithm in compressed sensing( CS)theory is applied to radar target recognition. First,an orthogonal matching pursuit( OMP) based on genetic algorithm( GA) is used to analyze the training samples and product taxonomic dictionaries quickly. Then,reconstruction errors of some testing samples are calculated so to recognize the targets. The simulations show that this algorithm has the advantages of conciseness,higher recognition rate and good robustness. Compared with some conventional methods,the proposed algorithm can increase recognition rate up to 20%.
出处
《电讯技术》
北大核心
2016年第1期20-25,共6页
Telecommunication Engineering
基金
国家自然科学基金资助项目(61571364)~~
关键词
雷达目标识别
高分辨距离像
压缩感知
稀疏表示
正交匹配追踪
遗传算法
radar target recognition
high resolution range profile
compressed sensing
sparse representation
orthogonal matching pursuit
genetic algorithm
