| 摘要: |
| 针对基于稀疏描述（SR）的识别算法的计算复杂度高，不利于算法实时、高效实现的问题，提出了一种快速稀疏描述（ESR）算法，以提高合成孔径雷达（SAR）图像目标型号识别的效率。考虑到SAR图像在一定的角度范围内惰性变化的特点，将每个型号目标的训练样本在一定方位区间内分别取平均，采用平均样本表征该方位区间内的若干个样本，以减少训练样本的数目，达到有效降低算法计算复杂度，提高SAR目标型号识别算法效率的目的。实测的MSTAR数据验证了所提快速算法的有效性。 |
| 关键词: 稀疏描述（SR） SAR图像 目标型号识别 计算复杂度 |
| DOI：10.3969/j.issn.1672-2337.2018.05.003 |
| 分类号:TN957 |
| 基金项目:国家自然科学基金（No.61701289，61601274，41471280） |
|
| Achieving Synthetic Aperture Radar Target Configuration Recognition via an Efficient Sparse Representation Based Algorithm |
|
LIU Ming， CHEN Shichao， LU Fugang， WU Jie， XING Mengdao
|
|
1.School of Computer Science，Shaanxi Normal University， Xi’an 710119， China;2.Xi’an Modern Control Technology Research Institute， Xi’an 710065， China;3.National Key Lab of Radar Signal Processing， Xidian University， Xian 710071， China
|
| Abstract: |
| The recognition algorithms based on sparse representation （SR） suffer from heavy computational complexity and thus are hard to realize in practical applications. To solve the problem，an efficient SR （ESR） algorithm is proposed in this paper. Taking account of the fact that SAR images have the property of varying slowly in a small range of azimuth angles，an average sample is calculated to represent the samples which cover related angles. The computational complexity can be dramatically reduced with a small number of the training samples. The efficiency of the proposed algorithm can be improved. The effectiveness of the proposed algorithm is verified on the measured MSTAR database. |
| Key words: sparse representation（SR） SAR image target configuration recognition computational complexity |
