Self-Calibration of Eye-to-Hand and Workspace for Mobile Service Robot (Heikkila, 1997), a 2-D pattern (Zhang, 2000), or a 1-D bar (Zhang, 2004). Once the camera parameters are obtained, features and objects can have their dimensional information based on 3-D computer vision. The hand-eye calibration is to compute the relative orientation and position between the robot arm (hand) and the camera (eye). Accord- ing to the arrangement of the camera, hand-eye systems are divided into two categories, which are the eye-in-hand and eye-to-hand configuration (Dornaika, 1998). The transformations in the two configurations are homogeneous when the camera can see the robot arm; hence, the most hand-eye calibration can be applied to both configurations. In essence, the hand-eye calibration problem can be solved by finding the transformation of the relative camera poses or relative hand-mounted reference poses. robot task and may vary from each other for differ- ent applications. Our idea to solve this problem is to use a laser distance sensor attached on a robot arm. Figure 2 shows an overall system in that a robot arm equipped with a laser distance sensor forms a calibration device. This device provides 3-D position measurements and its projected laser spots can be observed by cameras. Since the laser spots can be extracted and located in images when the laser spots are in the camera field of view, the camera pose relative to the robot arm can be determined based on the perspective geometry. This calibration scheme can also be applied when the field of view of each camera does not overlap. This offers a high flexibility for different tasks. Solutions and Recommendations Preliminaries Transformation of Cartesian Coordinates The rigid transformation from the camera frame to the robot base frame is determined by a rotation matrix R B , C and a translation vector T B , C . A point p C in the 3-D Cartesian coordinate system of the camera frame is transformed into the robot base coordinate system via: Figure 1. Camera networks for a mobile robot MAIN FOCUS OF THE CHAPTER Issues, Controversies, Problems A self-calibration scheme of the geometric struc- ture of a distributed vision system is essential to let an autonomous robot moving or manipulat- ing efficiently. We proposed a self-calibration scheme, which yields the complete geometric relationship between a robot arm, environmental cameras, and workspaces without any manual measurement. In some applications, cameras may not have overlap in the field of views. The existing methods (Svoboda, et al., 2005; Chen, et al., 2007) requires overlap region, which limit the flexibility of camera arrangement for a distributed vision system. Figure 1 shows an example in that the mobile robot needs to gather the information from the cameras and then to manipulate objects between tables. A complete system may require some cameras focus on small workspaces while others take wide-area surveillance to navigate the robot. The arrangement of cameras depends on the 231