Applications of Complex-Logarithmic Conformal Mapping in Still Image Coding

This paper evaluates the performance of an image compression system based upon the complex logarithmic conformal mapping. A staple in the robotics and visual systems fields, our attempts were to investigate the application of this algorithm in the field of image compression. The characteristics predisposed to the log sensor include a foveal region centered around the focal point which generates a high level of image reconstruction. In the extrafoveal region, we see a tapering off in resolution capacity as we approach the periphery of the sensor. This natural form of compression found in the human visual system is augmented by a spatial scaling constant a which, in addition to displacing the logarithmic singularity at the origin, directly regulates the size of the aforementioned foveal region. Two other characteristics govern our log mapping: the circumferential index κ and the visual field radius R. Comparisons between this space-variant sensor and a uniform resolution sensor yield further insight into the application of the log mapping.