基于改进ORB算法无重叠视野域的目标交接

Target handoff based on improved ORB algorithm on non-overlapping cameras

介绍了基于ORB(Oriented FAST and Rotated BRIEF)算法的无重叠视野域的目标交接。当目标离开摄像机1视野域时,利用ORB算法提取目标特征点,待匹配特征点通过PROSAC(Progressive Sample Consensus)算法剔除外点,保留有价值的特征点。当目标进入摄像机2视野域时,通过Vibe(Visual Background extractor)算法获取进入摄像机2的运动目标,并提取待确认运动目标的特征点,去除外点并与摄像机1视野域提取的特征点做匹配,匹配率达到阈值则表示目标一致,完成目标交接并进行跟踪。实验结果表明,文献提出的方案能够实现目标在多摄像机视野域间的快速交接。

This paper mainly introduce the use of ORB （Oriented FAST and Rotated BRIEF） algorithm on the non-overlapping cameras to handoff tracking target. When the target leaves camera 1, ORB algorithm is used to extract the feature points of the target. PROSAC （Sample Consensus PROSAC） algorithm is used to remove the outer points and persist valuable feature points. When the target is in the camera2, Vibe （Visual Background extractor） algorithm was used to get moving target which was into the camera2. We extract the characteristic points of the moving object that leave to be confirmed, removed outer points then matched the feature points extracted from Camera 1 and Cam- era2. If matching rate achieve the threshold that means tracking target is the same. Target is success- ful handoff and will be continue tracked in Camera2. Experimental results show that the scheme pro- posed in this paper, can handover target between the multi cameras in real time.