Content Based Image Synthesis

Abstract

A new method allowing for semantically guided image editing and synthesis is introduced. The editing process is made considerably easier and more powerful with our content-aware tool. We construct a database of image regions annotated with a carefully chosen vocabulary and utilize recent advances in texture synthesis algorithms to generate new and unique image regions from this database of material. These new regions are then seamlessly composited into a user's existing photograph. The goal is to empower the end user with the ability to edit existing photographs and synthesize new ones on a high semantic level. Plausible results are generated using a small prototype database and showcase some of the editing possibilities that such a system affords. © Springer-Verlag 2004.