摘要： 提出一种改进的运动目标检测算法，以准确检测不同光照条件下的运动目标。针对前景检测时出现的阴影，提出基于边缘信息的阴影消除算法。该算法与阴影方向无关，能去除目标各方向的大部分阴影，为视频监控系统的后续高级处理排除了阴影干扰。在配置为 2.0 GHz的P4计算机上运行，速度约为20帧/s。实验结果表明了算法的实时性、可靠性和准确性较好。

关键词: 目标检测, 背景模型, 阴影消除

Abstract: An improved moving object detection algorithm is proposed. It is robust to changes in background illumination and can detect multiple moving objects accurately in various complex scenes. Generally, the shadow is also detected as moving object because of the same characters between them. Another method of shadow removal based on edge information is presented too, which is independent of the direction of shadow, and can effectively remove the cast shadow of a moving object. Implemented by computer of P4 2.0 GHz, it runs at about 20 f/s. Experiments show that the method is reliable, and can be used in relevant real-time applications.

Key words: object detection, background model, shadow removal

中图分类号: 

• TP18