A Lucas-Kanade Moving Object Detection Algorithm Based on Gaussian Mixture Background Modeling

Release time:2024-08-09  Hits:

• Affiliation of Author(s):信息与控制工程学院
• Journal:Advanced Civil, Urban and Environmental Engineering
• Key Words:中文关键字：目标检测；混合高斯背景建模；Lucas-Kanade算法；局部约束，英文关键字：object detection；Gaussian mixture background model
• Abstract:Owing to the problems of computational time, poor real-time performance and inaccuracy of detection of inter frame difference method and most optical flow algorithms, a Lucas-Kanade moving object detection algorithm based on Gaussian mixture background modelling is proposed. Firstly use Gaussian mixture modelling estimate and update the background for input reference video frames, then apply Lucas-Kanade optical flow constraint algorithm calculate and compare the pixels in the area of video, finally find the changing gray region, and detect the object. Compared with the similar algorithms, the simulation results show that new algorithm can detect the moving object quickly and accurately, and has better robustness to the environment changes and target movement repeatedly.
• Note:王燕妮
• First Author:zhoujunni,wangyanni
• Indexed by:Journal paper
• Volume:卷：157
• Issue:期：1
• Page Number:页：97-106
• Translation or Not:no
• Date of Publication:2014-08-01