Recent image quality assessment (IQA) metrics achieve high correlation with human perception of image quality. Naturally, it is of interest to produce even better results. One promising method is to weight image quality measurements by visual importance. To this end, we describe three strategies - visual fixation-based weighting, quality-based weighting and weighting based on distribution of local quality scores about the mean. By contrast with some prior studies we find that these strategies can improve the correlations with subjective judgment significantly. We demonstrate improvements on the SSIM index in both its multi-scale and single-scale versions, using the LIVE database as a test-bed. / text

Inventory management and production control problem of a dealer operating in a decentralized spare parts network is analyzed in this dissertation. Spare parts network is assumed to be formed of two dealers and the problem of a dealer is considered under the assumption that the other dealer has a known policy. These dealers collaborate through inventory and service pooling. Furthermore, the dealers collaborate through sharing information on the net inventory status. Upon demand arrival, a dealer may request a part from the other dealer, in which case a payment is made. Under this competitive and collaborative environment, the optimal operating policy of an individual dealer is characterized under full information. Through computational analysis, the conditions under which the dealer under consideration is most profitable are identified. Finally, by comparing dierent pooling strategies and several information availability levels, the benefit of information sharing is quantified.