Visual tracking is a challenging problem, as an object may change its appearance due to pose variations, illumination changes, and occlusions. Many algorithms have been proposed t...
In this paper, a unified, segmentation-based approach is proposed to deal with both stereo reconstruction and moving objects detection problems using multiple stereo mosaics. Each...
- An architecture system and a method for tracking people are presented for sports applications. The system’s input is video data from static camera and the output is the real wo...
We present a new framework for robust 3D tracking, using a dynamic data driven coupling of continuous and discrete methods to overcome their limitations. Our method uses primarily ...
In this paper, we present an efficient algorithm for 3D object recognition in presence of clutter and occlusions in noisy, sparse and unsegmented range data. The method uses a robu...