A visual positioning and measurement system for robotic drilling

Abstract

In order to increase drilling efficiency, it is necessary for robotic drilling system to take advantage of offline programming method to drill a large number of holes automatically. Vision measurement is needed to correct work piece position because of inconformity between actual object and mathematical models. This paper introduces a visual positioning and measurement system for robotic drilling. A calibration board is designed to obtain hand-eye relationship and four laser range sensors are used to measure surface normal. This paper also summarizes workflow of offline programming and correction methods for different parts. Experiments give the result of hand-eye relationship and prove the position accuracy of vision measurement.