3D position/orientation measurement using model-based matching method and GA: proposal of calculation method and evaluation

This paper presents a vision-related technique for 3D position/orientation measurement. The proposed method utilizes the genetic algorithm (GA) and unprocessed grayscale image input from stereo-vision, in order to perform recognition of a target being imaged with known target object shape. In fact here, the problem to recognize the target shape and simultaneous detection of the position/orientation, is converted to an optimistic problem of a model-based evaluation function, named as surface-strips model-based fitness function that consists in the computation of the brightness difference between an internal surface and a contour-strips. In order to evaluate the proposed 3D recognition method, experiments by a robot manipulator equipped with two hand-eye cameras have been conducted to recognize a ball on a turning table. The results show its effectiveness of this method for real-time visual servoing by manipulator.