Camera motion and visual information fusion for 3D target tracking

This paper proposes a data fusion scheme for visual object identification and tracking by autonomous vehicles. In this scheme, image motion vectors fields, color features, visual disparity depth information and camera motion parameters are fused together to identify the target 3D visual and dynamic features. This paper also presents a detailed description of the 3D target tracking algorithm using an extended Kalman filter with a constant velocity dynamic model. Performance of the proposed scheme is discussed through experimental results.