兵工自动化2015，Vol.34Issue(5)：33-37,41,6.DOI:10.7690/bgzdh.2015.05.010
一种加速的PCA-L1增量子空间学习跟踪方法
Object Tracking Algorithm Based on Accelerated PCA-L1 of Incremental Subspace Learning
摘要
Abstract
For dealing with the large amount computation and not real time of increment subspace tracking algorithm, analyze computation bottleneck of L1-norm maximization target tracking algorithm, use and improve bounded particle re-sampling (BPR) algorithm, introduce accelerated proximal gradient (APG). Test and compare the tracking effect and speed with other classical methods. Experimental results show that, the method accelerate speed effectively, it can improve the real-time tracking, and has strong engineering application value.关键词
PCA-L1算法/l1范数/APG/粒子滤波Key words
PCA-L1 algorithm/l1 norm/APG/particle filter分类
信息技术与安全科学引用本文复制引用
王兵学,康林,黄自力..一种加速的PCA-L1增量子空间学习跟踪方法[J].兵工自动化,2015,34(5):33-37,41,6.基金项目
总装预研(402030203)基金资助项目 （402030203）
