摘要
主要关注基于核函数密度估计的Mean-Shift算法在视频序列图像中运动目标的识别与跟踪的应用。Mean-Shift算法避开全局搜索,因此达到了实时性的要求。本文对基于Mean-Shift理论的目标跟踪算法分析了其在视频跟踪中的优点,同时也指出了Mean-Shift算法的不足。
This paper mainly studies based on kernel density estimation theory of Mean-Shift in recognition and tracking of moving targets in video image sequences. Mean-Shift algorithm avoids the global search,so have a good to satisfy the requirement of real-time.This paper summarizes the probability density estimation theory,analyzes its advantages in video tracking,and at the same time points out the shortages of Mean-Shift algorithm.
出处
《自动化与仪器仪表》
2016年第4期20-22,共3页
Automation & Instrumentation
