筆觸參數於局部與大域之擾動控制與效應觀察

Abstract

在非擬真顯像鉛筆素描風格的呈現上，本論文提出了一種方法，用來將原始影像轉變成具有素描風格型態的影像，這個方法的主要的步驟可分成影像分群、區塊的分析及筆觸的生成。

首先利用影像簡化的技術，將原始影像切割成若干個區塊，然後分析每一個區塊的各種特性，例如：灰階值、面積大小和像素角度等。筆觸模型則會依據這些區塊的特性，去模擬鉛筆素描筆觸的生成。

透過不斷的實驗與觀察，去調整筆觸生成模型及各種參數所造成的影響，使得最終可以產生出令人滿意的素描影像圖。

The paper presents a method for converting original images to pencil sketch -styled images. Main steps in the process consist of image grouping, block analysis and stroke generate.

In the first step, the original image will be divided into several blocks by image simplification technology. The next is to analyze various imagery characteristics within each block, for instance, intensity, block size and pixel angle, etc. Stroke Model will then generates the pencil-sketch strokes based on the characteristics of these blocks.

Through a series of experiments and observations, the Stroke Model is turned and the effects are observed. Finally, the system can produce a satisfactory pencil-sketch-styled images.