We propose a new multi-target tracking approach, which is able to reliably track multiple objects even with poor\nsegmentation results due to noisy environments. The approach takes advantage of a new dual object model\ncombining 2D and 3D features through reliability measures. In order to obtain these 3D features, a new classifier\nassociates an object class label to each moving region (e.g. person, vehicle), a parallelepiped model and visual\nreliability measures of its attributes. These reliability measures allow to properly weight the contribution of noisy,\nerroneous or false data in order to better maintain the integrity of the object dynamics model. Then, a new multitarget\ntracking algorithm uses these object descriptions to generate tracking hypotheses about the objects moving\nin the scene. This tracking approach is able to manage many-to-many visual target correspondences. For achieving\nthis characteristic, the algorithm takes advantage of 3D models for merging dissociated visual evidence (moving\nregions) potentially corresponding to the same real object, according to previously obtained information. The\ntracking approach has been validated using video surveillance benchmarks publicly accessible. The obtained\nperformance is real time and the results are competitive compared with other tracking algorithms, with minimal\n(or null) reconfiguration effort between different videos
Loading....