基于混沌粒子滤波的视频目标跟踪

针对复杂场景中目标受光照、自身形变、遮挡等影响，本文以混沌粒子滤波为框架，建立多特征似然模型，进行目标遮挡处理，提出了一种鲁棒性好且抗遮挡性强的混沌粒子滤波跟踪算法。本算法中利用混沌优化搜索优化粒子，很好的克服了退化现象，减少了计算量，在多特征似然模型建立中，对特征选择做了改进，使算法鲁棒性更好，并在算法中添加了遮挡处理。理论数据及实际场景的仿真结果表明，本文提出的算法鲁棒性好且具有较强的抗遮挡能力。

Abstract

Referring to influence of illumination，shape transformation and occlusion on the tracking of the moving target in complex background, a anti-occluding and robust tracking algorithm is proposed based on chaos particle filter frame, in which the multi-cue likelihood model and occlusion dealing algorithm are constructed. In this algorithm, through chaos optimization search, the particles are optimized, the degeneracy problem is efficiently overcome and the computation is reduced. In the multi-cue likelihood model, some feature of choice is improved, which makes the algorithm robust. At the same time, the occlusion dealing makes the algorithm anti-occluding. Experimental simulation results based on the theoretical data and the actual scenes show that this algorithm is anti-occluding and robust.

参考文献

[1]

. CONDENSATION-Conditional Density Propagation for Visual Tracking[J]. International Journal of Computer Vision (S1573-1405), 1998, 1(29): 5-28.

[2]

. Data Fusion for Visual Tracking with Particles[J]. Proceedings of the IEEE (S0018-9219), 2004, 92(3): 495-513.

[3]

. An Adaptive Color-based Particle Filter[J]. Image and Vision Computing (S0262-8856), 2003, 21(1): 99-110.

[4]

. Hybrid Particle Filter and Mean Shift tracker with adaptive transition model[J]. Proc. of IEEE Signal Processing Society International Conference on Acoustics,Speech,and Signal Processing (S1520-6149), 2005, 2: 221-224.

[5]

. Sequential Monte Carlo tracking by fusing multiple cues in video sequences[J]. Image and Vision Computing (S0262-8856), 2007, 25(1): 1217-1227.

[6]

. Dim Target detection Based On Nonlinear Multi-Feature Fusion By Karhunen-Loeve Transfrom[J]. Optical Engineering (S0091-3286), 2004, 43(12): 2954-2958.

[7] . 基于多特征融合的图像匹配模式[J]. 强激光与粒子束, 2004, 16(3): 281-285.

强激光与粒子束|2004|16|3|281-285

. Image Matching Based On Multi-features Fusion[J]. High Power laser and Particle Beams, 2004, 16(3): 281-285.

[8] . 基于增强分形特征的人造目标检测[J]. 光电工程, 2006, 33(10): 45-47.

光电工程|2006|33|10|45-47

. Artificial target detection based on enhanced feature[J]. Opto-Electronic Engineering, 2006, 33(10): 45-47.

[9]

Li P，Chaumette F. Image Cues Fusion for Object Tracking Based on Particle Filter [C]∥ AMDO 2004，2004，3174：99-107.

[10] 光电工程|2003|30|4|15-19

. 一种弱小目标的自适应搜索策略[J]. 光电工程, 2003, 30(4): 15-19.

. An Adaptive searching method for faint targets[J]. Opto-Electronic Engineering, 2003, 30(4): 15-19.

[11]

Bergman N. Recursive Bayesian estimation：Navigation and tracking applications [D]. Sweden：Department of Electrical Engineering Link ping University，1999.

[12] 信息与控制|1998|27|3|232-235

. 利用混沌搜索全局最优的一种混合算法[J]. 信息与控制, 1998, 27(3): 232-235.

. A Hybrid Algorithm For Finding Global Minimum[J]. Control Theory And Applications, 1998, 27(3): 232-235.

刘世军, 彭真明, 赵书斌, 张启衡. 基于混沌粒子滤波的视频目标跟踪[J]. 光电工程, 2010, 37(7): 16. LIU Shi-jun, PENG Zhen-min, ZHAO Shu-bin, ZHANG Qi-heng. Tracking Target in Video Sequences Based on Chaos Particle Filter[J]. Opto-Electronic Engineering, 2010, 37(7): 16.

基于混沌粒子滤波的视频目标跟踪

关于本站 Cookie 的使用提示

全站搜索

基于混沌粒子滤波的视频目标跟踪

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索