Automatic analysis of eye tracker data

The movement of a persons eyes is an interesting factor to study in different research areas where attention is important, for example driving. In 2004 the Swedish national road and transport research institute (VTI) introduced Simulator III – their third generation of driving simulators. Inside Simulator III a camera based eye tracking system is installed that records the eye movements of the driver. To be useful, the raw data from the eye tracking system needs to be analyzed and concentrated into a number of measures relevant for the research at VTI. This thesis presents methods to analyze the data from the eye tracker and transform it into something more useful. A world coordinate system is setup to connect the eye tracking system with the real world in a consistent way. A set of measures is collected, mainly from ISO and SAE standards, to be used as output from the analysis. Finally an application is developed for performing the analysis. The application reads the data from the eye tracker and the simulator, analyzes the data, and outputs a set of eye movement measures usable for the researchers at VTI.