摘要
双目视觉技术是计算机视觉研究中的焦点,其能够在多种条件下获取目标的深度信息,实时交互性强,本文针对传统单目视觉目标检测与跟踪算法的鲁棒性差、无法适用复杂环境的弊端,结合双目视觉技术的工作原理,提出了一种融合灰度和视差的新型运动目标检测方法,并运用改进的多特征粒子滤波算法实现了运动目标跟踪,有效推进了人机交互体育运动的应用实现。
the binocular vision technology is the focus in the research of computer vision,it can acquire the depth information under various conditions,real-time interactivity.This paper aim the poor robustness and inability to apply in the complicated environments of traditional barrage visual,combined with the working principle of the binocular vision technology,to propose a new detection method of moving object base on the gray level and the disparity,and used an improved particle filtering algorithm to realize the moving target tracking,so as to promote the applications of human-computer interaction movement.
出处
《自动化与仪器仪表》
2016年第10期189-190,共2页
Automation & Instrumentation
关键词
双目视觉技术
运动捕捉
目标检测
跟踪算法
binocular vision technology
motion capture
target detection
tracking algorithm
