智能视频监控中目标跟踪算法的分析-通信与信息系统专业论文.docxVIP

AbstractWith the improving of the hardware equipment performance, intelligent video surveillance system is wildly used in the security monitoring field. Moving targets tracking is the significant module of which. In the case of targets fast movement, appearance and illumination changing and background disorder achieve the moving targets tracking is the key points and difficulties which the researchers concerned.In many targets tracking algorithms, our research concentrates on the more usual mean drift algorithm and particle filter algorithm in the intelligent video surveillance system area. Mean drift algorithm is an algorithm based on the nonparametric density function estimation, which convergences to the function peak along the biggest density function gradient change direction by the iteration method, such that get the position of targets tracking. Calculated amount of mean drift algorithm is less and easy to achieve it, which is commonly used in real-time targets tracking. But to the background disorder and targets small situation, the mean drift algorithm is more commonly lost the targets. The particle filter algorithm achieve the recursive Bayesian estimation of the target state through the monte carlo simulation method, such that extract the random state particles from the posterior probability of target state, and correct weight of each particles by observing the similarity between the target and template. Eventually in conjunction with each particle to represent target state distribution, and thus determine the target state. Particle filter algorithm represent the target state by statistic average of multiple particles, which still can get a more robust tracking result as the noise influence is more significant. But there exists drawbacks of more calculated amount and sample lack in the particle filter algorithm. This paper chapter 2 analyses theories and describes the algorithms in detail of above two algorithms.The main research contents and innovation i