Portraits Based On Feature Automatically Generates 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 developed an automatic feature-based portrait generation system, which can transform the input facial photo into a portrait in the style of a given artist's portrait. First, considering the MPEG-4 standard and some existed models, we propose a new model, which contains 64 feature points. Second, we extract the facial feature in some new ways. It is known that there are some fixed rules in laying facial features in a human face. Based on the statistical data, we divide the face into several small areas after detecting the iris, then extract corresponding feature points in each area using the fixed direction Active Contour Model. Finally, the system generates a portrait by field morphing, which uses an artist's finished work as a source image. The results show that the system is capable of producing portraits in different styles effectively.