Generating face ink portrait from face photograph

Abstract

The Chinese ink portrait requires sophisticated skills and the training for Chinese ink painting takes a long time. In this research, a Chinese portrait generation system is proposed to allow the user to convert face images to Chinese ink portraits. We search the image using Active Shape Model (ASM) and extract facial features from an input face image. As a result, a feature-preserved ink diffused image is generated. In order to produce a feature-preserved Chinese ink portrait, we use artistic ink brush strokes to enhance face contour constructed with the facial features. The generated portraits can be used to replace faces in an ink painting.