Reconstruction and analysis of human walking patterns using a computer 3D vision system

Human walking analysis is an important research area of biomedical engineering since it provides accurate information for medical rehabilitation procedures and design of rehabilitation equipment, medical diagnosis and orthopedics, pathological and aging evaluation, design of human prosthesis, and design of humanoid robots. In some applications, such as the design of prosthesis and rehabilitation systems in biomedical engineering, and equipment design and performance analysis in sports engineering, human walking reconstruction under several conditions is required in order to optimize the design. Human walking process is smooth and efficient but it varies from one person to another depending on age, height, gender, weight, health condition, and walking conditions. Most of the research work in the literature has been focused on the analysis of gait patterns of healthy and unhealthy people under normal walking conditions, and they use 2D reconstruction of human walking trajectories. The aim of this paper is to reconstruct and analyse human walking patterns of normal young adults under different gait conditions. A computer vision system to reconstruct 3D human walking trajectories is developed and presented in this paper. Several experiments with young adults walking under several conditions such as carrying a front load, carrying a lateral load, ascending, etc., are conducted. The results of these experiments have shown that human walking patterns vary according to the walking condition and therefore these variations should be considered in the design of prosthesis or rehabilitation systems.