Hamood-ur Rehman (University of Texas at Austin), "Artifact Assessment and Generation of Video Halftones" (2010)

Hamood-ur Rehman (University of Texas at Austin), "Artifact Assessment and Generation of Video Halftones". Advisor: Prof. Brian L. Evans (Dec. 2010)

With the advancement of high-resolution display technology, consumers expect high quality display of image and video data. In certain portable electronic books and handsets with micromirror displays, however, displays are binary. For a video stream, a video halftoning algorithm reduces the number of represented colors or gray levels to match the display, which results in visible spatial and temporal artifacts. Evaluating video quality vs. implementation tradeoffs in video halftoning algorithms is critical in display subsystem design. This dissertation concerns the design and analysis of binary video halftoning algorithms for display at 15-30 frames per second.  The dissertation proposes

1. an analysis framework for automated evaluation of two key temporal artifacts, flicker and dirty-window-effect, based on subjective testing;
2. design of new video halftoning algorithms with reduction in perceived temporal artifacts;
3. analysis of relative power consumption for video halftoning algorithms on bistable devices; and
4. design of enhancement algorithms for existing video halftoning algorithms to reduce flicker and/or dirty-window-effect temporal artifacts under both spatial and temporal quality constraints.

For details, please check here.