A Fuzzy Approach for Measuring Features Reliabilities in Particle Filtering Framework

In this paper, we propose a fuzzy inference system by which reliability of features can be measured. The reliability is defined as discriminative power of a feature in separating the target from background. We use particle filter because of its power and versatility in visual tracking. The reliability is measured based on observations diversity and spatial scattering of particles. Efficiency of our algorithm is demonstrated using color, edge and texture features. Experimental results over a set of real-world sequences show that our method''s performance is better than four common feature weighting methods.