Self-optimizing Visual Servoing Control for Microassembly Robotic Depth Motion

This paper presents a self-optimizing visual servoing control method for microassembly robotic depth motion. To measure micromanipulator depth motion, a normalized gray-variance focus measure operator is developed using depth from focus techniques. The extracted defocus features are theoretically distributed with one peak point which can be applied to locate the microscopic focal depth via self- optimizing control. Tracking differentiators are developed to suppress noises and track the features and their differential values without oscillation. Based on the differential defocus signals a coarse-to-fine self-optimizing controller is presented for micromanipulator to precisely locate focus depth. Experimental results of microassembly robotic depth motion demonstrate the performance of the proposed method with depth servo error of 7.5 mum.