基于DM8168的遗留物体检测算法设计

Design of abandoned object detection algorithm based on DM8168

针对智能监控中遗留物体检测算法存在物体之间的遮挡和将静止的人检测为遗留物造成的虚警问题,为了达到实际应用中低功耗、实时和系统稳定性的要求,设计了一种基于达芬奇DM8168平台的遗留物检测方案。该方案使用双重背景方法检测出由运动变成静止的目标,同时结合证据累加的方式解决了运动目标遮挡静止目标造成的虚警问题,并采用支持向量机分类器对静止的目标分为人和遗留物体,避免了将静止的人检测为遗留物的问题。实验结果验证了该算法可以达到预期的效果。

To solve the problems of blocking and false alarm caused by regarding the still people as remnants during aban-doned objects detection,and satisfy the requirements of low power consumption,real time and stability in practical application, an abandoned object detection scheme based on DM8168 was designed. In the approach,dual-background is used to detect the still objects coming from moving objects. It also solve the false alarm problem caused by the phenomenon that the moving objects block out the still objects in combination with evidence accumulating mode. Moreover,the SVM classifier is used to distinguish between people and abandoned objects to avoid the problem that the still people are regarded as abandoned objects. The experi-mental results show that the algorithm can achieve the desired effect.