Analysis of the kinematic variation of human gait under different walking conditions using computer vision

Abstract

This paper presents the analysis of human gait of three young adults under several walking conditions similar to those in real life. The aim is to evaluate how the human gait varies when different walking conditions are considered. In order to evaluate the human gait variations under different walking conditions, first a 3D computer vision system to reconstruct human gaits is developed and presented. Then, a set of kinematic gait parameters is defined in order to analyse the different gait patterns. Finally, several experiments are conducted with three normal young adults walking under different real-life conditions, such as ascending a slope walk, no arm-swing walk, carrying a front load walk, carrying a lateral load walk, fast walk and high-heel shoes walk. The results has shown that the human gait pattern vary according to the walking conditions, being the most common variation a slower walk with shorter steps than the normal walk. Since the aim of the paper is to identify and evaluate human gait variations under different walking conditions, only three subjects were considered in the study. However, in order to provide more precise and standard results the number of subjects must be increased considering different age, height, sex, weight, and health conditions of the participants. Human walking patterns are very important for medical diagnosis and orthopedics, pathological and aging evaluation, medical rehabilitation, and design of rehabilitation systems, human prosthesis and humanoid robots. However, most of the research work in the literature has been primarily focused on the analysis of gait patterns under normal walking conditions. This fact has led to a limited knowledge of the human gait patterns since real-life walking conditions are diverse, e.g. walking carrying a load. This paper presents the analysis of human gait of three young adults under walking conditions similar to those in real life. Human gait patterns vary with the walking conditions, which must be considered during the analysis, evaluation and diagnosis of gait performance, or during the design process of prostheses or rehabilitation systems.