中國水墨畫的模擬與合成

Abstract

水墨畫是一項歷史悠久的中國傳統藝術，但作畫所需的技法繁多且費時。在 畫作完成之前我們無法預知其將展現的風格為何。倘使畫家想嘗試另一種作 畫風格，他必須重新重複一次之前的過程。這種方式不但耗時而且不能確定 所得的結果確為自己所期望。在本篇論文中，我們提出自動模擬與合成水墨 畫的方法。經由一張現成的畫作，我們的方法自動找出其中的筆觸並且依照 使用者所想要的風格重新產生一張電腦合成的水墨畫。如此能不必經過實際 的作畫過程便能預視結果，不但節省時間而且使用者並不一定要熟稔水墨畫 技法便能完成結果。

Chinese Ink Painting is an art with long history in China. However, it requires so much time and painting techniques during the painting process. Besides, we will not know what the final results looked like before the paintings are finished. If a painter want to see the painting in different styles, he must repeat the entire painting process and still can not sure that the result is the one he wants. In this thesis, we propose a method to synthesize Chinese Ink Painting automatically. The algorithm extracts strokes from an existing Chinese Ink Painting image and re-paints it in other styles whatever we want. Using this method we can see resulting images without really performing the painting process. Moreover, the user even requires no painting technique to finish the job.