Vision based Manifolds for Surface Modeling in Robotic Demanufacturing

M. Bailey-Van Kuren (USA)

Keywords

Surface modeling, Demanufacturing, Machine vision,Manifolds, Topology

Abstract

Due to growing stream of end-of-life electronics, industry is starting to focus on the processing of electronic waste. In addressing problems associated with automated demanufacturing of electronic products, the need for a flexible product surface model was found in order to facilitate surface operations on the product. Consequently, a vision based manifold for surface modeling is developed. The vision based topology is proven to meet the definitions for a topological space and manifold. The application of the manifold to resolve product features is discussed. Furthermore, the model was implemented in a robotic work cell to test the feasibility of the model for path planning on the surface. It was found that a manifold based surface model provides a practical method to use machine vision feedback to construct a simplified representation of a complex product surface. A problem with the structured lighting approach was found and future directions to improve the applied lighting are discussed.

Important Links:



Go Back