ABSTRACT
This chapter emphasizes that even the global solution, which is based on combinatorial optimization techniques, can strongly benefit from a robotic interaction. It describes the machine vision algorithm for solution of large jigsaw puzzles, its experimental results, and the proposed integrative method using robotic assembly feedback. The chapter also describes the integrative robotic assembly system for solution of small jigsaw puzzles and its practical implementation by an IBM 7565 Cartesian robot. It focuses on the solution of the grasping and manipulation problems involved in fine assembly and in coping with uncertainty. The chapter states some future improvements and research directions. It discusses itegrative approach combining the application of machine vision techniques for local curve-matching estimation, mathematically sophisticated combinatorial optimization techniques to obtain global solution proposals, and robotic fine assembly techniques, which are used to verify these solutions and feed back partial results, thus enabling significant speedup in the iterative application of the previous modules.
