Model-based object tracking in monocular image sequences of road traffic scenes

摘要: Moving vehicles are detected and tracked automatically in monocular image sequences from road traffic scenes recorded by a stationary camera. In order to exploit the priori knowledge about shape motion of scenes, parameterized vehicle model is used for an intraframe matching process recursive estimator based on estimation. An interpretation cycle supports with state MAP-update step. Initial hypotheses generated using segmentation component which clusters coherently moving features into candidate representations images vehicle. The inclusion illumination allows taking shadow edges account during process. Only such elaborate combination various techniques has enabled us track under complex conditions over long (over 400 frames) sequences. Results real-world presented open problems as well future work outlined.