Human Gait Recognition Using Fuzzy Logic
In this paper, an efficient technique has been implemented for gait based human identification. This paper proposes a human identification system based on human gait signatures extracted using topological analysis and properties of body segments. The gait
- PDF / 282,935 Bytes
- 11 Pages / 439.37 x 666.142 pts Page_size
- 15 Downloads / 241 Views
Abstract In this paper, an efficient technique has been implemented for gait based human identification. This paper proposes a human identification system based on human gait signatures extracted using topological analysis and properties of body segments. The gait features extracted are height, hip, neck and knee of the human silhouette and a model-based feature i.e. area under hermite curve of hip and knees. The experimental phase has been conducted on the SOTON covariate database, which comprises of eleven subjects. The database also takes into account different factors that vary in terms of apparel, carrying objects etc. Subject classification is performed using fuzzy logic and compared against the nearest neighbor method. From the conducted experimental results, it can be accomplished that the stated approach is successful in human identification while some analysis prove that specific number of input variables and membership functions help to elevate the accuracy level. Keywords Gait
Hermite curve Fuzzy logic Nearest neighbor
P. Arora (&) S. Srivastava A. Chawla S. Singh Netaji Subhas Institute of Technology, New Delhi, India e-mail: [email protected] S. Srivastava e-mail: [email protected] A. Chawla e-mail: [email protected] S. Singh e-mail: [email protected] © Springer Science+Business Media Singapore 2016 M. Senthilkumar et al. (eds.), Computational Intelligence, Cyber Security and Computational Models, Advances in Intelligent Systems and Computing 412, DOI 10.1007/978-981-10-0251-9_27
277
278
P. Arora et al.
1 Introduction Human Gait is an image based human identification technique in which the pattern of synchronized and cyclic movement of the legs is used to recognize the person. It can be used to identify persons by using the video recording of the way a person walks. The gait analysis includes the extraction of some specific unique features of the persons under study to draw conclusions for the identification [1]. Human gait recognition can be performed without the subject being aware of the surveillance. However, the gait unique features may be affected by unwanted footwear, fatigue and injury as well. The continuous increase in the number of people has led to a need for security and surveillance. This has made gait analysis a recent area of research. Gait recognition has been classified into two types—model based and model free. Model based methods capture the human body configuration in motion. The features that are extracted during the motion are modeled and matched to the designed model features. It amalgamates the knowledge of the body shape and the kinematic gait features extracted from the model parameters. The main advantages of the model-based approach is that it can reliably handle occlusion (especially self-occlusion), noise, scale and rotation well, as opposed to silhouette-based approaches [2]. However, it creates many parameters from extracted gait features and hence resulting in a complex model. Due to the complexity involved in analysis of the model b
Data Loading...
