ABSTRACT
Motivated by the image space navigation function developed in, an off-line desired image trajectory generator is proposed based on a new image Jacobian-like matrix for the monocular, camera-in-hand problem. An adaptive, homography-based visual servo tracking controller is then developed to navigate the camera-in-hand pose along the desired trajectory despite the lack of depth information from a monocular camera system. The common image-based visual servoing algorithms do not have mechanisms to satisfy rigid body constraints which cause image-based visual servo robot systems to behave irregularly when the initial error is large. The common image-based visual servoing algorithms do not have mechanisms to satisfy rigid body constraints which cause image-based visual servo robot systems to behave irregularly when the initial error is large. The path planner and the tracking controller are analyzed through a Lyapunov-based analysis. A Lyapunov-based analysis is provided to support the claims for the path planner and to analyze the stability of the adaptive tracking controller.
