大谷淳,岸野文郎
遺伝的アルゴリズムを用いたマルチカメラ画像からの
人物の姿勢推定法の検討
Abstract:A new method for estimating human postures from multiple images using a Genetic Algorithm is proposed. In our algorithm, the posture parameters to be estimated are assigned to the genes of an individual in the population. For each individual, its fitness to the environment evaluates to what extent the human multiple images synthesized by deforming a 3D human model according to the values of the genes are registered to the real human multiple images. Natural selection based on the fitness chooses parents who generate children in the next generation, and for the new individuals, crossover and mutation are performed at random in the mating pool. After a certain number of repetitions for these processes, the estimated parameter values are obtained from the individual with the best fitness. Experiments using synthesized human multiple images show promising results for estimating 17 joint angle values for each degree of freedom of the joints and also the three translational and three rotational degrees of freedom of a human.