Angular estimation by image analysis and genetic algorithm

Abstract

In this paper, we propose a method which allows a precise estimation of the angular position of a non-deformable body using image processing and genetic algorithms. We present the geometric model used to describe the relationship between coordinates in the image of a mark situated on the non-deformable body and the angular position. We show that the estimation of an angular position is an optimization process. We describe the image processing step which allows one to detect the marks in the images. Finally, we show that the optimization process can be evaluated efficiently by a fuzzy genetic algorithm.