Seybold Report ISSN: 1533-9211
L. Mallika
Assistant Professor, Department of Electronics and Communication Engineering, Sridevi Women’s Engineering College, Hyderabad, India, mallikal123@gmail.com
Koonreddy vedasri
U.S Student, Department of Electronics and Communication Engineering, Sridevi Women’s Engineering College, Hyderabad, India
Patlolla Lavanya
U.S Student, Department of Electronics and Communication Engineering, Sridevi Women’s Engineering College, Hyderabad, India
S.Manasa
S.Manasa
Vol 17, No 07 ( 2022 ) | DOI: 10.5281/zenodo.6877853 | Licensing: CC 4.0 | Pg no: 201-207 | Published on: 25-07-2022
Abstract
Gait is a unique non-invasive biometric form it we can make the use to recognize persons, even when they prove to be unhelpful. Computer aided gait recognition systems generally use image sequences without considering covariates like clothing and possessions of carrier bags whilst on the move. Similarly, in gait recognition, there may exist unknown covariate conditions that may affect the training and testing conditions for a given every individual. A common techniques for gait recognition and measurement need a degree of intervention leading to the introduction of unknown covariate conditions, and hence this significantly limits the practical use of the present gait recognition and analysis system. To overcome these key issues, we propose a method of gait analysis for both known and unknown covariate conditions. For this purpose, we propose two methods A) Convolutional Neural Network (CNN) based gait recognition for known covariate conditions. This method can handle known covariate conditions efficiently B) Discriminative features-based classification method for unknown covariate conditions. It will focus on identifying and selecting unique covariate invariant features from the gallery and probe sequences. Here we can use some analysis like Local Binary Patterns (LBP), Histogram of Oriented Gradients (HOG), and Hara lick texture features. Furthermore, we will use another analysis is the Fisher Linear Discriminant Analysis for dimensionally reduction and selecting the most discriminant features. Gait recognition under strict unknown covariate conditions are three types namely Random Forest, Support vector machine (SVM), and Multilayer Perceptron. We evaluated our results using CASIA and OUR-ISIR datasets for both clothing and speed variations. As a result, we report that on average we obtain an accuracy of 90.32% for the CASIA dataset with unknown covariates and similarly performed excellently on the ISIR dataset. Therefore, our proposed method out performs existing methods for gait recognition under known and unknown covariate conditions.
Keywords: