基于二维小波变换的激光成像雷达目标识别算法

陈晓清; 马君国; 赵宏钟; 付强

doi:doi:10.3788/lop48.041001

激光与光电子学进展, 2011, 48 (4): 041001, 网络出版: 2011-02-22

基于二维小波变换的激光成像雷达目标识别算法

Target Recognition Algorithm Based on Two-Dimensional Wavelet Transform for Laser Imaging Radar

论文大纲

陈晓清 ^*马君国赵宏钟付强

作者单位

国防科技大学自动目标识别实验室, 湖南长沙 410073

图像处理目标识别二维小波变换奇异值 image processing target recognition two-dimensional wavelet transform singular value

摘要

提出了基于二维小波变换的激光成像雷达目标识别算法，首先对激光成像雷达目标的距离像进行二维小波变换；然后从近似分量和细节分量中提取奇异值特征，利用遗传算法对支持向量机参数进行智能优化；最后应用支持向量机对三种地面目标进行识别。仿真实验表明，该方法与直接应用距离像奇异值特征进行识别的方法相比，在高载噪比20 dB时的平均识别率提高了3.2%，在低载噪比10 dB时的平均识别率提高了6.5%，取得了比较好的识别效果。

Abstract

Target recognition algorithm based on two-dimensional (2D) wavelet transform for laser imaging radar is put forward. Firstly 2D wavelet transform of range image of laser imaging radar targets is made; then singular value feature is extracted from approximate part and detail part, gene algorithm is used to optimize parameters of support vector machine (SVM) intelligently; at last, SVM is used to recognize three kinds of ground targets. Simulation experiment shows that, compared with the method which directly uses singular value feature of range image to recognize targets, the average recogniton rate is raised by 3.2% at high carrier-noise radio (CNR) of 20 dB and average recogniton rate is raised by 6.5% at low CNR of 10 dB in this method, and good recognition is gained.

参考文献

[1] Douglas G. Youmans, George A. Hart. Three-dimensional template correlations for direct-detection laser-radar target recognition [C]. http://handle.dtic.mil/100.2/ADA390244, 1999

[2] Asuman E. Koksal, Jeffrey H. Shapiro, William M. Wells. Model-based object recognition using laser radar range imagery [C]. SPIE, 1999, 3718: 256～266

[3] Jianfeng Sun, Wei Lu, Qi Li et al.. Correlation target recognition for laser radar [C]. SPIE, 2006, 6027: 1～7

[4] 张贤达. 现代信号处理[M]. 北京：清华大学出版社，1995

Zhang Xianda. Modern Signal Process [M]. Beijing: Tsinghua University Express, 1995

[5] 瓦普尼克. 统计学习理论的本质[M]. 张学工译. 北京：清华大学出版社，2000

V. N. Vapnik. Essence of Statistic Learning Theory [M]. Zhang Xuegong Transl.. Beijing: Tsinghua University Express, 2000

[6] 马君国, 赵宏钟, 王微. 基于最优参数支持向量机的目标识别算法[C]. 全国模式识别学术会议，2009，40～44

Ma Junguo, Zhao Hongzhong, Wang Wei. Target recognition algorithm based on support vector machine of optimization parameters [C]. Pattern Recognition Conference of China, 2009, 40～44

陈晓清, 马君国, 赵宏钟, 付强. 基于二维小波变换的激光成像雷达目标识别算法[J]. 激光与光电子学进展, 2011, 48(4): 041001. Chen Xiaoqing, Ma Junguo, Zhao Hongzhong, Fu Qiang. Target Recognition Algorithm Based on Two-Dimensional Wavelet Transform for Laser Imaging Radar[J]. Laser & Optoelectronics Progress, 2011, 48(4): 041001.

基于二维小波变换的激光成像雷达目标识别算法

关于本站 Cookie 的使用提示

全站搜索

基于二维小波变换的激光成像雷达目标识别算法

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索