Skill recognition

摘要

This paper proposes a method for the recognition of skill in skiing from an image sequence. This method consists of two stages. The first stage estimates motion parameters, which are displacements of the pose of the skier, from the image sequence using a model-based trading method. The second stage evaluates some bases, which describes skill, from the motion parameters. Experimental results on real image sequences show that higher evaluated value corresponds to more skilfulness. This means that our computational approach can recognize skill like human intuition.