Hasan Al Hamada MSc Dissertation 2014/15

An Activity Recognition Application for iPhone

Supervised by F.Ciravegna

Abstract

In our modern world, the lack of activities diseases had become as deadly as smoking. Thus the demand for any possible solution is crucial. in this dissertation, a system will be built to encourage people to be more active by giving them a better representation for their daily activity. This project is aimed to develop an activity recognition application fo r iOS devices. In addition, the system classifies the user activity and represents it on a map to make it easy for the user to realise it. In order to achieve this aim, the application collects data from the device sensors and uses the iOS APIs to classify and analyses the data retieved from the device then send it to a server for further analysis. Moreover, to develop an application that uses the GPS, the developer needs to deal with the power consumption to get an efficient and powerful application in the same time which required a deep understanding for the knowledge related to the way the iOS control the hardware components of the device. 

However, the final product managed to control these two factors to get an application that work in efficient way without compromising the application functionality.

In order to get a better representation for the data on the map, an algorithm may be used to normalise this data, the algorithm that could be used to achive this result is the STRAVA slide algorithm.