We propose a natural scene statistic based quality assessment model Refer- enceless Image Spatial QUality Evaluator (RISQUE) which extracts marginal statistics of local normalized luminance signals and measures 'un-naturalness' of the distorted image based on measured deviation of them. We also model distribution of pairwise products of adjacent normalized luminance signals providing us with orientation distortion information. Although multi-scale, the model is defined in the space domain avoiding costly frequency or wavelet transforms. The frame work is simple, fast, human perception based and shown to perform statistically better than other proposed no reference algorithms and full reference structural similarity index(SSIM). / text

Low power, high speed serial transceivers are employed in a wide range of applications ranging from chip-to-chip, backplane, and optical interconnects. Apart from being capable of handling a wide range of data rates, the transceivers should have low power consumption (mW/Gbps) and be fully integrated. This work discusses enabling techniques to implement such transceivers. Specifically, three designs: (1) a 0.5-4 Gbps serial link which uses current recycling to reduce power dissipation and (2) a 0.5-2.5 Gbps reference-less clock and data recovery circuit which uses a novel frequency detector to achieve unlimited acquisition range and (3) a 2-4 Gbps low power receiver architecture capable of resolving multiple signalling formats with a simplified XOR based phase rotating PLL will be presented. All the three circuit topologies are highly digital and aim to address the requirements of wide operating range, low power dissipation while being fully integrated. Measured results obtained from the prototypes illustrate the effectiveness of the proposed design techniques. / Graduation date: 2012 / Access restricted to the OSU Community at author's request from Dec. 2, 2011 - June 2, 2012