Using Salient Lenses for Adaptive Image Display
A Project for Topics in Mobile and Pervasive Computing
Mobile computing devices vary widely in terms of their display capabilities. It is important that screen "real estate" be used effectively in order to improve the user experience and usability of the device (for example, maximizing text readability, or the sizes of selectable UI widgets). This course project will explore the application of salient lenses - a somewhat generic term that refers to the techniques and algorithms which maximize the viewable area of regions considered salient (or important).
Within the scope of a course project, I will specifically be looking at raster graphic images as both input and output. It will be assumed that any target device will possess a raster graphic display. The objective of the course project will be to explore existing approaches to improve image saliency, while considering the inherent computational cost of each. The salient lens, when operating client-side, should scale within a target device's computational power. Server-side operation of the salient lens will also be considered.
Quantifiable experiment results will come from testing the performance of various techniques - either by porting the implementations into a form executable on a mobile device, or by simulation on a desktop machine. Subjective results may also be obtained by user study, such as the user's opinion of whether the difference in presentation of output and improvement in usability is worth the amount of any additional latency.