一种基于背景自适应的运动目标检测与跟踪算法

背景图像差分法是运动目标实时检测中常用的方法,但缺乏背景图像随监视场景光照变化而及时更新的合理方法,限制了该方法的适应性。对此,文章首先提出了一种自适应背景更新方法;然后利用最大类间方差法实现运动目标的自适应阈值分割,并利用基于形态学方法的连通区检测算法检测运动目标;最后以Kalman滤波为运动模型实现对运动目标的连续跟踪。实验结果表明:所提方法可随着光照条件的变化,实时、准确地检测出运动目标并实现稳定跟踪。

A Moving Object Detection and Tracking Method Based on Adaptive Background Image

The method based on background image difference is commonly used for real-time detection of moving object. However, no reasonable method has been designed for automatic background updating along with the illumination variance, which limits its applications. To overcome the problem, a new self-adaptive background updating algorithm was first presented in this paper, and then self-adaptive threshold segmentation of moving objects was implemented using the Otsu method, and moving objects were detected by the binary connected component analysis using the morphological method, and Katman filtering was used for objects tracking. Experimental results demonstrated that the proposed method could detect and track moving objects exactly and quickly along with the variance of illumination.