作者: Subramanian Sridharan , Simon Denman , Vinod Chandran
DOI:
关键词:
摘要: Person tracking systems to date have either relied on motion detection or optical flow as a basis for person detection and tracking. As yet, not been developed that utilise both these techniques. We propose a person system uses both, made possible by novel hybrid flow-motion detection technique we developed. This provides the system with two methods of detection, helping to avoid missed detections the need predict position, which can lead errors in mistakes when handling occlusion situations. Our results show that our is able track people accurately, with an average error less than four pixels, system outperforms current CAVIAR benchmark system.