Walking pattern analysis using GAIT cycles and silhouettes for clinical applications

Abstract

In recent years biometrics play a vital role in recognizing the person and authentication. Recent studies prove that the gait cycle is unique for every individual. Gait refers to the walking pattern of an individual. Gait cycle is calculated by the right toe on the same right toe on period. Human Gait cycle and the angles calculated from head-to-toe portions of a person are important measures for both habitual and clinical analysis purposes. In most cases these parameters would furnish sufficient details for further deep learning and analysis serving as medical clues. These parametersprovide useful and validated results for medical rehabilitation. This paper proposes evaluation of Gait cycle for abnormal walking pattern of different diseases. The graphs were plotted for Gait cycle versus knee flexion angle and gait cycle versus hip to ankle angle variations which provide sufficient information for classifying normal and abnormal walk pattern. The main idea proposed here is to classify the different walk patterns without the involvement of medical tools. This proposed work will be helpful in obtaining the required and initial clinical details of persons by means of their gait without any direct medical investigation on them and facilitates in classifying different abnormal walk patterns.