基于粒子滤波和Mean-shift的自适应目标跟踪方法

目标跟踪技术一直是计算机视觉的核心内容。本文结合粒子滤波与Mean-shift跟踪方法，提出了一种新的自适应目标跟踪方法，通过利用粒子滤波获取目标的初始位置，进而采用Mean-shift跟踪方法，实现目标跟踪的准确定位，同时，通过抑制背景特征分布，更新目标特征分布，从而在跟踪过程中自适应调整目标的模板表示。实验结果表明了本文提出方法的有效性。

Adaptive Target Tracking Algorithm Based on Particle Filtering and Mean-shift

Target tracking is an important topic in computer vision. In this paper, a new adaptive target tracking algorithm based on Particle Filtering and Mean-shift was proposed. The initial position of target was acquired by using Particle Filter, then the accurate position of target was finally obtained by using the Mean-shift method. At the same time, the target template representation was adaptive adjusted through the background characteristics suppression and the target characteristics updating. Experimental results demonstrate the effectiveness of the method.