Visual Servoing by Means of Structured Light: Visual-based Robot Task Optimisation from Controlled Light Pattern Projection

Visual servoing is a popular technique for performing robotic tasks like relative positioning or target tracking. Classic visual servoing assumes that visual features can be extracted from camera images of the target object or scene. However, what happens when the target object is uniform? Can a robot position itself with respect to a white wall or avoiding a non-textured object using just visual feedback? What happens when the scenario is complex or a priori unknown like a region of the sea floor? In cases like these, visual servoing assumptions fail. This book studies the use of structured light to provide visual features. A rigorous comparison of the most relevant structured light patterns is provided. Then, structured light is integrated to a visual servoing framework. This book shows that a suitable design of the structured light pattern is able to provide robust and unambiguous visual features. Furthermore, the control law can be optimised in order to obtain nice properties like decoupling and robustness against image noise and calibration errors.

Jordi Pagès, PhD: graduated in Computer Science in the University of Girona, Catalunya, Spain, in 2001. He received the Ph.D in Information Technologies from the University of Girona and the University of Rennes, France, in 2005. He is currently a research engineer at Davantis Technologies, Barcelona.