Efficient tracking with AdaBoost and particle filter under complicated background

Abstract

Particle filter, which is the probability technique, can be used for the robust tracking to the noise and the occlusion. However, when many objects are tracked simultaneously, the real-time tracking becomes difficult as the computational cost increases. While, the AdaBoost has an ability that it has the remarkable efficiency as a statistical technique in pattern recognition. AdaBoost can be used to detect an object region for the efficient tracking with a particle filter. However, it is difficult to detect the moving object under the complicated background by AdaBoost. This paper proposes an improvement of efficiency of particle filter by introducing further distinction features using AdaBoost for the complicated background. © 2008 Springer-Verlag Berlin Heidelberg.