Global Tone: using tone to draw in Pen-and-Ink illustration

摘要

We present a new illustrating algorithm of global tone, enlightened by the skill called artistic tone. It is structure-aware and computationally driven by "visibility", which refers to quantitatively measuring how visible a point (or a region) on a mesh is within a virtual camera space. The feature lines are sketched by silhouettes and/or suggestive contours, which then are enhanced by colorful inks, mimicking tonal values. To overcome that computing tone throughout surfaces is prohibitive, the global shape descriptor of Gaussian visibility is proposed, which is fast and robust such that the rendering pipeline is finished in real-time. The line drawings are largely improved as the descriptor enables our approach to convey more shape cues beyond shades. We demonstrated the plausibility of global tone with various models, showing that our experimental results are comparable to or better than state-of-the-art.