基于功率谱三值化的联合变换相关目标识别方法  

Joint Transform Correlation for Target Recognition Based on Ternary-valued Power Spectrum

作　　者：李源[1] 杨丽丽[2] 陈惠连[1] 罗雪梅[1] 

机构地区：[1]中国工程物理研究院电子工程研究所,四川绵阳621900 [2]青岛农业大学电子信息学院,山东青岛266109

出　　处：《探测与控制学报》2009年第3期51-55,共5页Journal of Detection & Control

基　　金：国防预研项目资助(4260110-5)

摘　　要：为了提高目标识别的准确性和精确性,提出基于功率谱三值化的联合变换相关目标识别方法,从算法上对联合变换功率谱进行了修正,将联合变换功率谱进行了三值化(+1,0,-1)处理。与二进制联合变换相关器比较,三进制联合变换相关器大大提高了联合变换相关性能和分辨率,对输入目标敏感,能更好地鉴别目标。仿真实验结果证实了这种方法的可行性,较已有的相关识别方法在判断目标是否相同时更有效,提高了目标识别的准确性,算法也简洁快速。In order to improve the veracity and accuracy of target recognition, a new method of joint transform correlation for target recognition is proposed which is a new power spectrum ternary-valued method. The joint transform power spectrum has been modified in arithmetic and has been ternary valued （positive one, zero, negative one）. Compared with binary-valued joint transform correlator, ternary-valued joint transform correlator can better differentiate targets because it is greatly improved in performance and discrimination of joint transform correlation and it is sensitive to input target. The simulation results prove that this method is feasible. Compared with the existing correlation recognition method, the proposed method is more effective to judge whether targets is the same,its discrimination of target recognition has been improved,and moreover, its arithmetic is succinct and fast.

关 键 词：联合变换相关 二值化非线性调制 三值化非线性调制 联合功率谱 目标识别 图像识别 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]