Simulating realistic procedural Chinese-style brushwork by mimicking hand movement and adjusting brush model automatically

Abstract

Many non-photorealistic rendering systems produce illustration-like images from 3D geometry, but they seldom synthesize the style of Chinese ink painting. Our intention is to develop a rendering system that generates images in Chinese painting style based on the trail of hand movement while executing the brush. This system can automatically simulate Chinese brush strokes without the needs of users’ knowledge about Chinese paintings. Chinese painting masters usually paint with their impressions; thus, they only draw the features that are relevant to the painting objects. They also concern a lot on the manipulation of brush strokes and black ink. Mimicking the drawing habits of Chinese painters and simulating the hand movements of brush execution are the main contribution of our research. We present an algorithm that analyzes different reference images created from the 3D input data, and determines the shape of the strokes that should be applied, including the “stroke head”, “stroke tail” and with varying darkness and thickness. We will then demonstrate the results of using these novel algorithms on different 3D models.