Markerless human motion capture by Markov random field and dynamic graph cuts with color constraints

Abstract

Currently, many vision-based motion capture systems require passive markers attached to key locations on the human body. However, such systems are intrusive with limited application. The algorithm that we use for human motion capture in this paper is based on Markov random field (MRF) and dynamic graph cuts. It takes full account of the impact of 3D reconstruction error and integrates human motion capture and 3D reconstruction into MRF-MAP framework. For more accurate and robust performance, we extend our algorithm by incorporating color constraints into the pose estimation process. The advantages of incorporating color constraints are demonstrated by experimental results on several video sequences.human motion capture, Markov random field, dynamic graph-cut, color constrains

An inexpensive solution for motion analysis

“…Different methods have been developed to measure human motion including goniometers, 17 accelerometers, 15,18–20 inertia-based and electromagnetic sensors, 8,10,21–23 active and passive optical motion capture systems 4–6,9,13,14 and even markerless optical motion capture devices, 24–26 and ultrasound. 27 Initially Richards 9 compared different optical motion capture systems, followed by Mündermann et al, 25 who presented the evolution of methods for motion capture.…”

Segmentação e reconhecimento de gestos em tempo real com câmeras e aceleração gráfica

“…Quando não se usam marcadores, o problema de localizar os membros do usuário se torna bem mais difícil. Muitos autores tentam contornar esse problema com o uso de várias câmeras [70,21,76,23,24,54]. Quanto mais câmeras, há mais informação que pode ser usada para se recuperar a posição dos membros do usuário.…”