This paper presents a new algorithm to track mobile objects in differentscene conditions. The main idea of the proposed tracker includes estimation,multi-features similarity measures and trajectory filtering. A feature set(distance, area, shape ratio, color histogram) is defined for each trackedobject to search for the best matching object. Its best matching object and itsstate estimated by the Kalman filter are combined to update position and sizeof the tracked object. However, the mobile object trajectories are usuallyfragmented because of occlusions and misdetections. Therefore, we also proposea trajectory filtering, named global tracker, aims at removing the noisytrajectories and fusing the fragmented trajectories belonging to a same mobileobject. The method has been tested with five videos of different sceneconditions. Three of them are provided by the ETISEO benchmarking project(http://www-sop.inria.fr/orion/ETISEO) in which the proposed trackerperformance has been compared with other seven tracking algorithms. Theadvantages of our approach over the existing state of the art ones are: (i) noprior knowledge information is required (e.g. no calibration and no contextualmodels are needed), (ii) the tracker is more reliable by combining multiplefeature similarities, (iii) the tracker can perform in different sceneconditions: single/several mobile objects, weak/strong illumination,indoor/outdoor scenes, (iv) a trajectory filtering is defined and applied toimprove the tracker performance, (v) the tracker performance outperforms manyalgorithms of the state of the art.
展开▼
