A Robust Part based Method for Human Gait Recognition
Manjunatha Guru V G1, Kamalesh V N2
1Manjunatha Guru V G*, VTU RRC, Visvesvaraya Technological University, Belagavi, Karnataka, India.
2Kamalesh V N, Professor & Dean, T-John Institute of Technology (TJIT),Bangalore, Karnataka, India.
Manuscript received on March 30, 2020. | Revised Manuscript received on April 05, 2020. | Manuscript published on April 30, 2020. | PP: 1006-1009 | Volume-9 Issue-4, April 2020. | Retrieval Number: D7657049420/2020©BEIESP | DOI: 10.35940/ijeat.D7657.049420
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: This paper explored a new part based gait recognition method to address the gait covariate factors. Firstly, three robust parts such as vertical-half, head, and lower leg are cropped from the Gait Energy Image (GEI). Since, these selected parts are not affected by the major gait covariates than other parts. Then, Radon transform is applied to each selected part. Next, standard deviations are computed for the specified radial lines (i.e. angles) such as 0 0 , 300 , 600 , 900 , 1200 and 1500 , since these radial lines cover the horizontal, vertical and diagonal directions. Lastly, fuse the features of three parts at feature level. Finally, Support Vector Machine (SVM) classifier is used for the classification procedure. The considerable amount of experimental trails are conducted on standard gait datasets and also, the correct classification rates (CCR) have shown that our proposed part based representation is robust in the presence of gait covariates.
Keywords: Covariate, Gait, Part based, Radon.