Volumetric cloud generation using a Chinese brush calligraphy style

Clouds are an important feature of any real or simulated environment in which the sky is visible. Their amorphous, ever-changing and illuminated features make the sky vivid and beautiful. However, these features increase both the complexity of real time rendering and modelling. It is difficult to design and build volumetric clouds in an easy and intuitive way, particularly if the interface is intended for artists rather than programmers. We propose a novel modelling system motivated by an ancient painting style, Chinese Landscape Painting, to address this problem. With the use of only one brush and one colour, an artist can paint a vivid and detailed landscape efficiently. In this research, we develop three emulations of a Chinese brush: a skeleton-based brush, a 2D texture footprint and a dynamic 3D footprint, all driven by the motion and pressure of a stylus pen. We propose a hybrid mapping to generate both the body and surface of volumetric clouds from the brush footprints. Our interface integrates these components along with 3D canvas control and GPU-based volumetric rendering into an interactive cloud modelling system. Our cloud modelling system is able to create various types of clouds occurring in nature. User tests indicate that our brush calligraphy approach is preferred to conventional volumetric cloud modelling and that it produces convincing 3D cloud formations in an intuitive and interactive fashion. While traditional modelling systems focus on surface generation of 3D objects, our brush calligraphy technique constructs the interior structure. This forms the basis of a new modelling style for objects with amorphous shape.

Grantor dc:publisher.institution

Department of Computer Science

Year dc:date.issued

2014

Author dc:creator

• Wei, Chen

Advisors dc:contributor.advisor

• Gain, James
• Marais, Patrick

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/11427/13034

OAI identifier oai:identifier

oai:open.uct.ac.za:11427/13034