Feature-Based Automatic Portrait Generation System

Creating a portrait in the style of a particular artistic tradition or a particular artist is a difficult and also an interesting problem in non-photorealistic rendering. This paper develops an automatic feature-based portrait generation system, which can transform the input facial photo into a portrait in the style of a given artistpsilas finished work. First, we propose a new face model, which contains 64 feature points. Second, we extract the facial features using rules and geometric information. Based on the statistics, we divide the face into several small regions, each of which only contains one feature. The fixed-direction open active contour models are employed to extract corresponding feature points in each region. Finally, the system generates a portrait by feature-based morphing, which uses the artistpsilas work as the source image. The results show that the system is capable of producing portraits in different styles effectively.