The ability of humans to analyze, capture, and recall visual information significantly outperforms their ability to analyze and recall other types of sensory information. This fact makes humans rely heavily on their visual sense for extracting information and learning about their surrounding, and for planning and executing purposeful acts. As a result, the need to reliably process and transmit visual data has become central to many applications. In particular, with the rapid development and migration of end-user applications and services towards transmission media which place high constraints on bandwidth, such as the Internet and wireless media, there is a growing need for the development of new efficient image and video compression techniques, which offer reduction in bit-rates and improvement in quality at low bit-rates. Although the importance of exploiting human perception has long been recognized within the signal and image processing community, the previous research efforts in image and video compression have concentrated on developing methods to minimize not perceptual but rather mathematically tractable, easy to measure, distortion criteria. While non- perceptual distortion measures were found to be reasonably reliable for higher bit-rates (high quality applications), they do not correlate well with the perceived quality at lower bit-rates and they fail to guarantee preservation of important perceptual qualities in the reconstructed images despite the potential for a good signal-to-noise ratio (SNR). Our research interest is in developing adaptive perceptual-based image and video coding systems, which discriminate between signal components based on their perceptual relevance for achieving robustness and increased performance in terms of quality and bit-rate. Achieving this objective requires having reliable tools for predicting the sensitivity of the human visual system to distortions. The main objectives of this research can be summarized as follows: 1 . Investigate the spatio-temporal masking properties of the human visual system, and develop reliable techniques to adaptively measure the masking characteristics and perceptual relevancy of the different components contained in the visual source material.2. Derive a reliable tractable perceptual distortion metric for assessing visual quality at a local or global level. 3. Investigate adaptive quantizer designs based on the derived perceptual distortion metric.4. Develop and investigate robust perceptually-based lossless and lossy image and video coding techniques, with emphasis on audiovisual applications disseminated through highly constrained-bandwith communication systems. The research program is complemented by the education plan, in which a Signal and Image Processing Outreach Program is being developed to stimulate the interest of middle and high school students and undergraduates in engineering and science. Another main component of the educational plan is to integrate the research part into the educational process by incorporating some of the research activities in undergraduate and graduate courses as well as in undergraduate-oriented research projects, and by introducing new courses related to the research effort. Another goal is to research efficient ways to make use of technology in education.
