Realizing Anti-Artistic Filters by Image Analogies

Abstract

Artistic filters are commonly used to generate artistic images from photographic ones. In this article, we present a realistic-image synthesis system that works contrary to the traditional artistic filters by taking advantage of an emerging technique, called "image analogies", for modifying an input image with filtering effects specified by a pair of example images. Also paying attention to the high potentials of the present technique, we examine several effective applications with attempts at producing test images.