Title :
Automatic Target Recognition Based on HRRP Using SKO-KPCA
Author :
Zhengwei Zhu ; Jianjiang Zhou
Author_Institution :
Coll. of Inf. Sci. & Technol., Nanjing Univ. of Aeronaut. & Astronaut., Nanjing, China
Abstract :
In this paper, an adaptive and data-dependent single kernel optimization (SKO) algorithm is developed to improve the performance of radar target feature extraction and recognition by optimizing the kernel function of iterative kernel principal component analysis (KPCA). Based on SKO-KPCA and support vector machine (SVM), a radar target high resolution range profile (HRRP) feature extraction and recognition approach is proposed, and ensures, while comparing with other approaches, the satisfactory performances which are illustrated through automatic target recognition (ATR) experiments of Su-27, F-16 and M2000.
Keywords :
feature extraction; image recognition; image resolution; iterative methods; optimisation; principal component analysis; radar computing; radar resolution; radar target recognition; support vector machines; ATR; HRRP; SKO-KPCA; SVM; adaptive single kernel optimization; automatic radar target recognition; data-dependent single kernel optimization; high resolution range profile; iterative kernel principal component analysis; radar target feature extraction; support vector machine; Educational institutions; Feature extraction; Iterative algorithms; Kernel; Principal component analysis; Radar scattering; Space technology; Support vector machines; Target recognition; Testing; Automatic target recognition (ATR); High resolution range profile (HRRP); Kernel principal component analysis (KPCA); Single/fusion kernel optimization (SKO/FKO); Support vector machine (SVM);
Conference_Titel :
Computational Sciences and Optimization, 2009. CSO 2009. International Joint Conference on
Conference_Location :
Sanya, Hainan
Print_ISBN :
978-0-7695-3605-7
DOI :
10.1109/CSO.2009.89
