Research on the application of rectangle object constraint in active vision of mobile robot

In this paper, the method of indoor mobile robot's visual localization is studied. In an active vision system, the robot equipped with a camera can be controlled and regulated in two degree of freedom platform, the camera optical axis always points to the center of the target to ensure mobile robot can be precisely positioned in large field motion. In this paper, a method based on rectangular object constraint is used to measure the pose and position of the camera, and then the pose and position of robot itself were calculated by using the translation parameters of the camera and yaw and pitch angles of platform feedback, precise pose and position of robot will be gotten. Experimental results verify the effectiveness of the method. What the authors do is solving the FOV problem and applying rectangular target constraint location method in active vision system.