Human Identification through Gait Recognition

Aliya Amirzhanova

Abstract


Nowadays there are several image based human biometrics, such as iris recognition, fingerprints, face. However, all of them require close human contact. To solve this problem there is a relatively new technique, called gait recognition, which can be done at a distance. Moreover, it has advantages such as recognition of low resolution videos, recognition when individual information is confidential [6]. Certainly, there are effects of covariates such as change in viewing angle, change in shoe, walking surface, carrying conditions, and elapsed time [8] that make gait recognition problem more challenging for research. There are several approaches for studying gait recognition system such as model-based and model-free one. This project work was based on model-free approach. The main aim is to apply Machine Learning techniques to the gait recognition application system and find an optimal solution for the identification, while the dataset is robust to covariates.

Full Text:

PDF

Refbacks

  • There are currently no refbacks.


Copyright (c) 2014 Aliya Amirzhanova