自适应背景下运动目标阴影检测算法研究

Shadow Detecting for Moving Objects Based on Self-adaptive Background

智能视频监控系统的主要难点就是如何检测出运动目标的阴影并将他去除。用一种改进的自适应背景检测算法准确检测运动物体的位置和形状。然后根据阴影的颜色变化、结构等特点,分别采用基于RGB彩色模型和基于HSV彩色模型的阴影检测法检测阴影。在仿真实验中,对两种方法进行定量和定性的分析,在阴影检测率、识别率、复杂度和实时性等方面做出了比较。结果表明,两种方法都有较强的适应性,具有良好的阴影检测效果。根据各自的优缺点,可应用在不同领域中。

One of the main difficulties in intelligent video surveillance is how to detect and suppress the shadow in the scene, The moving objects detection algorithm uses an improved background subtraction algorithm which is based on self--adaptive background to detect the position and the shape of objects. Then according to the shadow＇s characters such as the color variation and the structure,two shadow detection algorithms which respectively based on the RGB color model and the HSV color model are introduced. In experiment,the two shadow detection algorithms are analyzed on the shadow detection accuracy, the shadow discrimination accuracy,the complexity and the real - time capability, and made the comparison of them. The resuits of experiment show that the shadow detection algorithms are effective and robust, and can be applied on different fields according to their advantages and disadvantages.