Camera system quality behavioral model using modulation functions and human visual pathway for imaging applications.

摘要

There has been an increasing interest in the development of imaging system quality behavioral model that can effectively predict the perceptual quality of captured images and video streams. After decades of research work, this quality assessment is still an active and evolving research topic. Examples of some quality assessment in pictorial segments include clarity in details, representation of color, lossy compression and pixel artifacts. These characteristics form the basis for defining sharpness, graininess, tone, scale, and color rendition, respectively. This thesis aims at discussing some of the open challenges and difficulties currently encountered in the Imaging industry for developing a good quality assessment approaches, and providing a vision for future directions. This is done by presenting a behavioral model which combines the physical parameters of the image forming system and retinal properties of the human visual pathway. Proposed model is validated by using two 1.3M camera module systems under carefully standardized viewing conditions which are detailed in this thesis. Valid conclusions are proposed for wider usage of the proposed model in the imaging industry.