Estimation of Gait Parameters from 3D Pose for Elderly Care

Abstract

For elderly people, walking, standing up from a chair, turning and leaning are necessary for independent mobility. These mobilities such as gait depends on a complex interplay of major parts of the nervous, musculoskeletal and cardiorespiratory systems. Every individuals gait pattern is influenced by age, personality, mood, sociocultural factors and predominantly the persons health condition. In order to understand the health condition of an elderly person, analysis of gait patterns became an important aspect. Gait parameters such as cadence, step length, step duration etc. analyzed out of gait patterns proved as an important factor in estimation of the healthy daily living. For this purpose, gait data of several elderly individuals is collected many times over a period of time using Kinect sensor. The acquired data consist of RGB image sequences and depth data. From this data, 3D pose of the individual is identified. These 3D poses are used to extract the necessary gait parameters of the individual. The extracted gait parameters will be used in future to assess the health condition of the individual.