Eye tracking for low vision aids: toward guiding of gaze

Eye tracking technique in the visibility study of low vision was newly introduced in the previous report, where we examined the ease of finding public signs on the streets and in the interior of buildings by low vision people. We got a conclusion that they hardly notice these signs. In this report we continue our research in this direction. We describe details of eye tracking technology applied to low vision. We devise calibration method for low vision. We describe analysis of eye tracking data on the basis of simplified gaze circle model of sight of low vision, leading to a conclusion that it is possible as well for low vision to locate regions of interest (ROI) by applying classical method of scanpath analysis. We also show a preliminary result of public sign recognition in the view by using a fast pattern matching technology called "boosting," linking to a future system for guiding the gaze of low vision to a missing public sign and zooming into it.