Gait recognition under non-standard circumstances

Abstract

This thesis will look at the biometric feature gait, and how different circumstances impact the ability to perform authentication. We will use an accelerometer sensor in order to record the data. First of all are we going to see if it possible to recognize users under different circumstances, then we will look for a common pattern among these different circumstances. We will also briefly look into whether people walk in the same way given the same circumstances. In order to answer these questions will we perform an experiment and do a thorough analysis of the data obtained. In a world where e.g. mobile devices becomes more and more important, it is obvious that the protection of these devices must be satisfactory. A possible security feature of such devices is to use gait authentication. When this shall become a reality however it is important that the authentication works under circumstances which are common during a normal day.