Distributed Architecture for Intelligent Robotic Assembly Part III: Design of the Invariant Object Recognition System

Abstract

In previous chapter it has been described the overall architecture for multimodal learning in the robotic assembly domain (Lopez-Juarez & Rios Cabrera, 2006). The acquisition of assembly skills by robots is greatly supported by the effective use of contact force sensing and object recognition. In this chapter we will describe the robot’s ability to invariantly recognise assembly parts at different scale, rotation and orientation within the work space. The chapter shows a methodology for on-line recognition and classification of pieces in robotic assembly tasks and its application into an intelligent manufacturing cell. The performance of industrial robots working in unstructured environments can be improved using visual perception and learning techniques. In this sense, the described technique for object recognition is accomplished using an Artificial Neural Network (ANN) architecture which receives a descriptive vector called CFDP vision systems as a sensorial mode for robots have a growing demand requiring more complex and faster image processing functions in order to implement more sophisticated industrial applications, like assembly automation.