计算机技术与发展2018，Vol.28Issue(6)：106-109,4.DOI:10.3969/j.issn.1673-629X.2018.06.024
基于联合直方图的自适应粒子滤波跟踪算法
An Adaptive Particle Filtering and Tracking Algorithm Based on Joint Histogram
摘要
Abstract
In order to solve the problem that traditional particle filter algorithm tracking the target is not ideal due to its appearance chan-ging and partial occlusion,we propose an adaptive particle filtering algorithm based on joint histogram of color and texture. This algorithm selects the global feature ( color histogram) and local feature ( LBP texture) to combine color and LBP texture histogram. By calculating the Bhattacharyya distance of the current target combined histogram and the initial target combined histogram as the basis of particle up-date weight,it can not only select an optimal set of particles but also keep its diversity with a smaller number of particles then,which solves the problem of particle degradation and the shortage of particles in re-sampling process in the traditional particle filter algorithm. Experiment shows that compared with the traditional particle filter algorithm,the proposed algorithm can robustly track the moving target with changes of the appearance or partly occluded.关键词
目标跟踪/粒子滤波/颜色特征/纹理特征Key words
tracking target/particle filter/color histogram/texture histogram分类
信息技术与安全科学引用本文复制引用
刘少林,朱文球,孙文静,王业祥..基于联合直方图的自适应粒子滤波跟踪算法[J].计算机技术与发展,2018,28(6):106-109,4.基金项目
湖南省重点研发计划项目(2016RS2020) （2016RS2020）
