Gait recognition is a distinct biometric technique for identifying people at a distance based on their walking style. It is used extensively in video surveillance, medical, forensics, digital security, and defense applications. However, the effectiveness of gait recognition can be adversely affected by several covariates such as clothing and the carrying conditions of an individual. Furthermore, identifying a person from multiple view-angles is another significant and complex challenge in gait recognition. Therefore, an appropriate selection of gait features is a challenging task in human gait recognition. To select an optimal subset of gait features, we proposed a novel adaptive firefly algorithm (AFA) based on the flashing patterns and behavior of fireflies. In this paper, gait energy image (GEI) is used as a gait template to extract the essential features from the input gait data. This proposed algorithm is applied to these gait templates to choose the most relevant and informative gait features. Finally, these selected features are subjected to principal component analysis (PCA), and multi-class linear discriminant analysis (LDA), to improve gait recognition performance. Extensive experiments are done on the CASIA-B dataset have revealed that the suggested approach significantly improves the effectiveness of gait recognition than current state-of-the-art approaches.
An Adaptive Firefly Optimization Algorithm for Human Gait Recognition
Advances in Sustainability sci. & technol.
Smart and Sustainable Technologies: Rural and Tribal Development Using IoT and Cloud Computing ; Chapter : 28 ; 305-316
2022-07-28
12 pages
Article/Chapter (Book)
Electronic Resource
English
Firefly algorithm for discrete optimization problems: A survey
Online Contents | 2017
|Firefly algorithm for discrete optimization problems: A survey
Springer Verlag | 2017
|Improved Human Gait Recognition
British Library Conference Proceedings | 2015
|Method for scheduling UAVs based on chaotic adaptive firefly algorithm
European Patent Office | 2020
|