Fast space-variant image analysis through steerable Gabor-like filters

The log-polar image geometry, first introduced to model the space-variant topology of the human retina receptors in relation to the data compression it achieves, has become popular in the active vision community for the important algorithmic benefits it provides. Despite these advantages, foveated sensing has not been widely used due to the lack of specific image processing tools. We demonstrate that it is possible to perform multichannel space-variant image processing with high computational efficiency through Gabor-like steerable kernels in the cortical domain.