Low-complexity fuzzy relational clustering algorithms for Web mining
IEEE Transactions on Fuzzy Systems
Hi-index | 0.00 |
In this paper, we exploit motion segmentation to enhance the robustness of a particle filtering based tracking process. We first propagate hypotheses from particle filtering to blobs of similar motion to target to achieve a more accurate prediction of regions of interest in the state space. This makes a new importance sampling scheme. After having identified the moving target, a representative model is learnt from its spatial support. This model is integrated as a reference in the next correction step of the tracking process. Hence, the proposed particle filter combines both motion and color information in an original way. It improves the performance of particle filtering in complex situations of occlusions compared to a simple Bootstrap approach as shown by our experiments on real fish tank sequences.