CAJ | 학술논문

基于合成孔径雷达图像的目标识别技术在军事、民用等领域都具有十分重要的作用。针对SAR（ Synthetic Aperture Ra-dar）图像在像素域稀疏表示识别算法中存在的高维问题，在分析其图像统计特性的基础上，提出一种基于压缩感知的合成孔径雷达图像目标识别方法。该方法利用扩展最大平均相关高度滤波器训练样本，生成模板，提取模板广义二维主分量特征构成过完备字典，求解测试样本在字典下的稀疏表示系数，根据系数能量特征完成分类识别。对MSTAR数据库中合成孔径雷达图像进行仿真实验，结果表明，该方法复杂度低，识别时间短，是一种可行且有效的合成孔径雷达图像目标识别方法。
기우합성공경뢰체도상적목표식별기술재군사、민용등영역도구유십분중요적작용。침대SAR（ Synthetic Aperture Ra-dar）도상재상소역희소표시식별산법중존재적고유문제，재분석기도상통계특성적기출상，제출일충기우압축감지적합성공경뢰체도상목표식별방법。해방법이용확전최대평균상관고도려파기훈련양본，생성모판，제취모판엄의이유주분량특정구성과완비자전，구해측시양본재자전하적희소표시계수，근거계수능량특정완성분류식별。대MSTAR수거고중합성공경뢰체도상진행방진실험，결과표명，해방법복잡도저，식별시간단，시일충가행차유효적합성공경뢰체도상목표식별방법。
Target recognition in SAR images plays an important role in military , civil and other areas .A method based on compressed sensing is presented for SAR target recognition after analysing the statistical characteristic of SAR images in order to solve the high dimensional problem of SAR image in pixel domain with sparse representation recognition algorithm .The method trains the samples and generates templates using the extended maximum average correlation height filter , extracts the template ’ s generalised two-dimensional principal component features to form an over-complete dictionary , the sparse representation coefficient of the test sample ’ s feature is computed base on the optimal dictionary .Classification and recognition are realised according to the energy feature of coefficient .Simulation experiment is carried out based on SAR images in MSTAR database , results show that the proposed method has lower complexity and short recognition time, it is a feasible and effective method for SAR images target recognition .