作者: Matthew Field , Zengxi Pan , David Stirling , Fazel Naghdy
DOI: 10.1108/01439911111106372
关键词: Group method of data handling 、 Mixture model 、 Motion capture 、 Cluster analysis 、 Motion estimation 、 Data mining 、 Body movement 、 Hidden Markov model 、 Artificial intelligence 、 Motion control 、 Computer science
摘要: Purpose – The purpose of this paper is to provide a review various motion capture technologies and discuss the methods for handling captured data in applications related robotics.Design/methodology/approach approach taken compare features limitations trackers common use. After introducing technology, summary given robotic‐related work undertaken with sensors strengths different approaches are discussed. Each comparison presented table. Results from author's experimentation an inertial system discussed based on clustering segmentation techniques.Findings trend methodology towards stochastic machine learning techniques such as hidden Markov model or Gaussian mixture model, their extensions hierarchical forms non‐linear dimension reduction. resulting empirical models tend handle uncertainty well suitable incrementally updating mo...