Determination of the significance of visual image difference features
This topic is a sub-project of the Emmy-Noether Group for Perceptually Optimal Reproduction of Color Images considering Device Limits. For that purpose hypotheses-based image difference features will be systematically analyzed in a visual experiment and statistically evaluated. The extracted visual image difference features will be arranged according to their visual significance and integrated in an image difference measure.
This image difference measure will be used to optimize the image reproduction considering the physical device limits of printers, displays and projectors. The image difference measure shall serve as an objective function of an optimization problem with the device limits as constraints. Numerical methods will be developed and implemented in a performance optimized software. The aim is to solve such high-dimensional problems in reasonable time for multispectral images.