A New Method of Correspondence for Multiple Cameras Based on Texture Energy

This paper proposes a new target matching method for multiple cameras based on texture energy. The target region is represented by texture energy instead of traditional color features. In our method, the frame-difference is used in multi-target detecting, and the kalman filter is utilized in multi-target tracking. None camera calibrations are required in the new method, neither the constraint that all objects stand in the same plane. The main features of the method are highlighted as follows: (1) Non geometry constraints are required, i.e. none the posture and the position of the target is required, and neither the position of camera is required. (2) Many types of objects including cars and persons are supported. (3) It is robust to changes of the light intensity. Experimental results show that the method is more efficient than the region-sift descriptor based method of correspondence, because the dimension is far less than the region-sift based method. However, the method in this paper is less accurate than the region-sift descriptor based method.