Hand Tracking Based on Improved Particle Filters with Elliptical Region Covariance Descriptors

Abstract

In the process of human computer interaction, hand tracking is of great importance. A practical hand tracking method based on improved particle filters with elliptical region covariance descriptors is proposed. Firstly, an elliptical tracking window containing the hand is determined manually in the initial frame. Based on the HSV color model, the color feature of bare hands is extracted, and color histograms of the target model and the candidate model are obtained. Then the observation likelihood function can be determined. A first-order system equation is used as the motion model. In order to take into account rotation changes of hands, an elliptical region covariance descriptor is used as the target feature model. The particle number threshold is preset, and the particle impoverishment can be improved by resampling method. Experimental results demonstrate that the proposed method can track the moving hand accurately. The proposed hand tracking method can be used in the fields of human computer interaction and augmented reality.