Subject Identification Using Walking Posture

Abstract

Human posture can provide important information through non-verbal communication. In recent years, studies of human posture recognition are becoming more and more numerous in the areas of personal healthcare, environmental awareness, human-computer interaction and surveillance systems. This paper proposes a simple method of posture recognition during walking using the leg inclination. The developed system uses mobile phone tilt sensors and a deflection sensor. Mobile phones are positioned on the back of the foot above and below the knee and the Flex sensor is positioned on the foot. The purpose of this system is to identify the subject after the walking posture. The test analysis is conducted on four subjects of different height and mass. The results are compared using MATLAB processing algorithms.