Research on Video Moving Object Detection Based on Improved Gaussian Mixture Model
Jun Xie, Yingnan Liu, Yongqi Zhang, Zhi‐Qiang Wang · 2023
Moving object detection is one of the most important and difficult problems in computer vision. How to find and track objects from video images and extract useful information timely and accurately is an urgent problem to be solved. A video moving object detection algorithm based on improved Gaussian mixture model is proposed. The algorithm uses multiple Gaussian distributions to model the background and uses the characteristics of the model to optimize the number of Gaussian mixture model sub-models and background threshold, so as to realize the moving object detection in complex external environment. Experimental results show that the algorithm can better detect the edge details of moving objects in complex environments and complex background changes and can effectively and accurately extract moving objects in videos.