Object Recognition Mobile Robot Guidance
Having presented an embodied vision approach to visually-guided robot navigation, this chapter discusses putting it into practice. The ability to uniquely identify and locate objects in its environment enables a robot to differentiate similar objects in order to interact with a specified object. This chapter describes a method for object recognition and navigation that is designed specifically for robot interaction. It is based on canonical-views, a novel object representation which realises principles of embodied vision. This chapter also presents a direct algebraic method for finding the orientation and position of the specified object.
KeywordsMobile Robot Object Recognition Object Representation Robot Motion Robot Navigation
Unable to display preview. Download preview PDF.