Date of Publication :8th August 2024

Abstract:Gait recognition methodology used to recognize the person by their walking style. This technique is performed at a distance as video without the cooperation of the person who has to be recognized. Gait recognition have been used in various applications like access control, surveillance and forensics. Gait recognition strategies can be divided into three different techniques based on the setting as machine vision method, floor sensor method and wearable sensor method. In machine vision method, one or more cameras will be used for video capturing for processing the image transformation. This method is also termed as Gait Energy Image since the gait features were extracted from the blurring silhouette image got from the video. In floor sensor method, features of gait will be extracted from the images generated by the sensing floor. Wearable sensor method uses the techniques like gait kinematics, gait kinetics, and electromyography for extracting the gait signatures. Gait recognition is liable to variations, such as view angle, clothing, walking speed, shoes and carrying status. Convolution neural network is used to extract the high level features from less quality video by the following steps: preprocessing the video to obtain the input data. Extracting the features and comparing with pre trained neural network for recognition. A human acknowledgment and recognizable proof is viewed these days as an essential field of research. The most unique parts of human are the ear, odor, heartbeat, voice, the iris, periocular portion of eye, fingerprint, gait, sweat, face, etc,. Without the human interaction to identify a person is quite challenging with low resolution images. Gait recognition is one of the biometric technology which can be used to identify people without their knowledge. The proposed system uses Deep Convolutional Neural Network to extract the gait features of a person by training the neural network architecture with Gait Energy Image.

Reference :