视频监控中运动目标快速检测方法研究

摘要

提出了一种背景差分的改进算法.通过像素灰度值归类与形态学相结合的方法来提高背景重构的效率,进一步提高系统的实时性.实验结果表明,改进的算法能够有效克服光线变化、雨雪天气等环境噪声的干扰,适应高密度运动目标检测场景,进一步提高检测速度和效果.%An improved algorithm of background subtraction is presented. Using the method of combining classification of pixel gray and morphology to improve the efficiency of background reconstruction can further improve the systems real-time performance. Experimental results show that the improved algorithm can effectively overcome the light changes, swinging leaves and other environmental noise interference, and it can adapt to the high density moving target detection scene and further improve the detection speed and effectiveness.