摘要： 介绍了一种基于TMS320C67xx实现的便携式实时视频检测系统设计解决方案,深入研究了CMU大学的运动检测算法和形态学开运算滤波等方法,根据实际需要采用了一种综合帧间差、自适应背景相减和形态学滤波的运动检测方法,仿真结果证明了此方法的快速性、有效性及对光照变化、噪声影响的鲁棒性。最后给出了此检测方法在基于DSP的视频检测系统中的简单实现流程,并探讨了今后的应用前景。

关键词: 便携式视频检测系统, 形态学开运算滤波器, 背景自适应, TMS320C67xx

Abstract: This paper presents a portable system designed for video monitoring based on TMS320C67xx,and takes a deep study on CMU’s moving object detection algorithm and the operator of morphology open operation filter.It gives out a moving object detection method,which combines frame difference,adaptive background subtraction and morphology filter.Some experiments prove the method given in this paper is fast,effective and robust to effect of illumination and noise.At the end of this paper,an example using this method is given,which is DSP-based video monitoring system,and its application prospect is also discussed.

中图分类号: 

• TP274.4