With the more and more complex applications, multisensor integration and fusion have become a new research area. In this paper, the current researches of the multisensor fusion are discussed as one of the popular fields. The problems in the research are summarized, including information description space, data association and synchronization, prior information and fusion structure, etc. Then the fusion method is discussed and classified into two categories: probability based and artifical intelligence based methods. Finally the importance of fusion structure is emphasized in the field of multisensor fusion and its future research is outlined.

This paper proposed a new way to extract sample points, which can be applied to the robot hand-eye coplanar calibration. Under the structure of global fixed camera, the positions of these sample points in image plane coordinates as well as in the robot Cartesian coordinates can be derived with this way correspondingly by using image processing and coordinates mapping. This approach developed the method of corner detecting and improved the precision of image processing. The experiment shows its efficiency.