Font Size: a A A

Moving Object Detection And Tracking In Intelligent Video Surveillance System

Posted on:2016-10-12Degree:MasterType:Thesis
Country:ChinaCandidate:G L SuFull Text:PDF
GTID:2348330461480196Subject:Pattern Recognition and Intelligent Systems
Abstract/Summary:PDF Full Text Request
Intelligent Visual Surveillance technology has been a great hot research area in computer vision, which uses visual computing technology and video image processing to realize automatic surveillance by image detection, tracking, analysis and understanding. Target tracking, an important task of intelligent visual surveillance computer vision, is widely used in human motion recognition, video surveillance, video retrieval aspects, virtual reality, human-computer interfaces, etc. In this paper, the based on particle filter moving target tracking algorithms is analyzed comprehensively and its improvement is proposed. The main research contents are presented as follows:Firstly, due to the poor tracking performance in complicated scenarios, multi-feature fusion based on particle filter algorithm is proposed. This algorithm exploits the fusion feature of color feature and local binary pattern (LBP) texture feature under the framework of particle filter to improve the robustness of target tracking. In order to get the accurate color model of target, the multi-part color histogram with spatial information is introduced. Comparing with the particle filter tracking algorithm based on single feature, experimental results show that the proposed method is more robust in terms of pose changes and illumination changes, especially in scenarios where the target object contains cluttered background with similar color distributions.Secondly, a new particle filer (SPF) based on the Bacteria Foraging Optimization (BFO) algorithm is proposed to deal with the particle degeneracy problem,. The proposed method take the sample weight as the fitness function, through a set of operations, including chemotaxis, reproduction, and elimination and dispersal operation, the optimal samples are found. The simulation results show that the proposed BFO-based particle filter (BFO-PF) outperforms the Standard particle filter (SPF), the Auxiliary Particle filter (APF) and the Genetic algorithm-based particle filter (GA-PF).Finally, in order to verify the validity of improved particle filter based on bacteria foraging optimization, the BFO-PF algorithm is applied in the actual video target tracking, and compared with the based SPF tracking algorithm and the based GA-PF tracking algorithm. The target tracking results show that the algorithm proposed in this paper uses fewer particles can achieve accurate target tracking. Furthermore, when the target was in the complicated situation, like repeatedly occlusion or larger postural change, the algorithm can maintain tracking accuracy.
Keywords/Search Tags:Intelligent visual surveillance, Moving target tracking, Particle filter, Bacteria foraging optimization resampling
PDF Full Text Request
Related items