摘要
针对运动目标在运动过程中的交叉、遮挡等情况,采用自适应阈值的Vibe算法来压缩背景杂波和相关噪声,进而对运动目标进行检测。采用基于Camshift优化的粒子滤波算法对运动目标进行跟踪,该算法在粒子滤波算法的基础上结合Camshift算法的优点,加入当前观测信息,使粒子更好地采样于目标周围,提高了粒子效率,节省了算法时间。实验表明,自适应阈值的Vibe算法能够准确检测复杂场景中的运动目标,并能够适应噪声干扰和光照变化,而基于Camshift优化的粒子滤波算法能够在目标快速运动、遮挡情况下对目标进行准确跟踪。
An improved algorithm in the image of video detection and tracking for the moving target is proposed in the course of the campaign cross, shelter, etc. The Vibe algorithm with adaptive threshold compress clutter and associated noise in the background are used to detect detect the moving target. The particle filter algorithm based on Camshift algorithm tracks the moving target, the algorithm combines the advantages of Camshift algorithm based on particle filter algorithm, adding the current observation information to make better sampling of the target particles around, improving particle efficien- cy and saving the algorithm time. Experimental results show that the improved Vibe algorithm can accurately detect the mov- ing target in the complex scene and have the ability to adapt to changes in noise and light, and the particle filter algorithm based on Camshift algorithm can accurately track the moving target in the actual scene.
出处
《计算机与数字工程》
2015年第4期576-581,590,共7页
Computer & Digital Engineering
