作者: E.B. Meier , F. Ade
关键词:
摘要: The detection of objects in every frame a sequence is often not sufficient for scene interpretation. Tracking can increase the robustness, especially when occlusions occur or temporarily disappear. In this paper we present stochastic tracking approach which based on CONDENSATION algorithm (conditional density propagation over time) that capable multiple with hypotheses range images. A probability function describing likely state propagated time using dynamic model. measurements influence and allow incorporation new into scheme. Additionally, representation fixed number samples ensures constant running per iteration step. Results data from different sources are shown automotive applications.