Survey of data association of moving objects tracking in video sensors network

Object tracking is a crucial task in computer vision systems for surveillance, traffic monitoring or intelligent homes. In all these cases, tracking is based on association of observations. Conventional tracking approaches assume similarity in space, time and appearance of objects in successive observations. However, Observations of objects are often widely separated in time and space when viewed from multiple non overlapping cameras. Moreover, appearance of one object is dramatically different among cameras due to the differences in illumination, pose and camera parameters. On the other hand, real world´s time-space constraint (motion path information)on moving object and sensors position are always available in advance. Thus, this paper presents a survey of data association method across visual sensors network for moving objects tracking. In particular, previous research on various data association methods of non overlapping cameras is reviewed.