2P2-O08 筋活動度データベースと統計的運動認識を利用した筋張力推定(感覚・運動・計測)

概要

論文の詳細を見る
Somatosensory informations such as muscle tension are essential for analyzing human's motion, but it is difficult to measure those by appearance. So human's musculoskeletal model and estimation of those information have been researched. The typical method is to formulate a numerical optimization problem, and we can also estimate biologically valid muscle tensions by adding electromyography (EMG) information to the problem. But using EMG data is not realistic because EMG sensor is generally not available. So in this paper, we propose a method to estimate biologically valid muscle tensions without EMG sensors by database having EMG information in every motions. We classify the database by motion pattern to avoid human body's redundancy problem. Then we estimate only by observed motion information and the appropriate pattern's database. We demonstrate this approach can estimate tensions which are close to the tensions estimated with EMG.
一般社団法人日本機械学会の論文
2011-05-26