Picture Quality Improvement Schemes
for HDTV Applications

Hiroshi Shimamoto
Kunal Ghosh
EE 362 / Psych 221
March 20, 2006

Introduction

In this project, we further investigate the picture quality improvement schemes that are enabled by the availability of the high resolution imager. Specifically, we simulate and verify the picture quality improvement achieved using a 4x2 color filter array (CFA). We analyze its advantages over the standard Bayer CFA and suggest directions for future work to overcome some of its limitations. Next, assuming that we are limited in spatial resolution by the CFA, we explore new demosaicing algorithms that seek to recover the resolution through intelligent colour interpolation. We present a preliminary approach to such demosaicing algorithms, examine its advantages, and suggest future work to overcome its limitations. Our project is therefore a continuation of the work in the picture quality improvement , and with the ultimate goal being a commercially feasible single-sensor HDTV camera system that is not only inexpensive and portable, but also further enhances the picture quality that is achievable using expensive and bulky present-day systems.

View the full report as Adobe PDF format.
View the presentation file as Microsoft Power Point format.
Link to resources in Matlab and Appendixes.
Download the zipped folder containing all the files and data associated with the project.