Effect of Carrying Load on Gait Recognition using J48 with Knee Joint Movements
Jyoti Rana1, Nidhi Arora2

1Dr. Jyoti Rana, Department of Computer Science, Naran Lala College, Navsari, India. 

2Dr. Nidhi Arora, SoluSoft Technologies Pvt. Ltd., Ahmedabad, India.

Manuscript received on 25 April 2020 | Revised Manuscript received on 07 May 2020 | Manuscript Published on 22 May 2020 | PP: 16-20 | Volume-9 Issue-7S July 2020 | Retrieval Number: 100.1/ijitee.G10060597S20 | DOI: 10.35940/ijitee.G1006.0597S20

Open Access | Editorial and Publishing Policies | Cite | Zenodo | Indexing and Abstracting
© 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: Background: Gait patterns are influenced by various factors. Every person walks differently in different scenarios and it becomes difficult to identify the person correctly by his walking style. Research question: Is it possible to correctly recognize a person through his gait while he walks carrying load? What is the effect of distance on a person’s gait while he walks with carrying load? Objective: The paper is an attempt to study the effect of load on gait of a person. As the knee angle varies from person to person in varying conditions, we have studied the effect of load on knee angle and thus on gait recognition. Methods: Experimentation is done on 41 subjects of age group 18-30 years carrying bilateral weights of 2.5kg in both hands. Data collected from accelerometer has been studied using J48 decision tree classification algorithm. Results: Results of experiments shows 86.12% accuracy in recognizing subjects carrying load up to a distance of 60 meters on a flat surface. The FAR and FRR are found to be 0.77% and 27.52% respectively. Conclusion: Carrying load affects the gait of the subject. This makes it difficult to recognize the subject while he walks carrying load. After walking for some distance, the gait pattern and knee angle of the subject shows significant variations.

Keywords: Gait, FAR, FRR, EER, ROC.
Scope of the Article: Image Processing and Pattern Recognition