以Web-mining為基礎的3D色票視覺化系統

不論是專業的畫家或設計師、還是一般的使用者或初學者，顏色選擇在很多場合都是一門非常重要的課題。不管選擇的是鮮豔的色彩，或是柔和的色彩，或是冷色系的色彩，我們最終的目的莫過於讓觀看者，可以感受到色彩的美妙之感；進而留下感動的深刻印象。色票是顏色設計的基礎，它可以讓使用者在混亂且眾多的顏色當中，避免產生視覺誤差。依據其所標示的色相、彩度及明度，使用者可以精確地決定適合的顏色，並在大量色票中快速找到目標。然而，現在的色票大多都採平面設計，只能觀察二維變化，無法完整透視三維的色立體並且直覺地選擇顏色。本研究將開發出一種色彩搜尋方式，提供使用者多一種應用的工具。色彩心理學(color psychology)是一門由色彩學以及心理學所共同衍生出的學科，不同的顏色透過色彩與心理的關聯性，給予觀者不同的認知。它提供了一種實用且直覺的對應連繫著色彩與人們的感覺。在本研究的系統中，我們將建立一套依照使用者需求，所量身打造出的色票系統。使用者先在3D情緒座標中直覺點出當前的感覺，或者選擇符合的關鍵字眼例如藝術家的名字，然後便利用web mining軟體在網路上搜索相關的圖片，分析這些圖片，把其中的顏色抽取統整出相符合的色票資訊，最後再把這些色票資訊架構成3D視覺化系統，並允許使用者可以觀察、控制系統所擷取出的3D色票，並提供對照參考。

關鍵字

色票；色立體；網路資料探勘；色彩心理學

並列摘要

In this paper, we propose an emotion-based visualization system for 3D color charts. Choosing colors is an important issue not only for professional artists and designers, but also for amateurs and beginner groups. By carefully arranging the tints and shades of colors, an artist allows viewers to feel what the artist is feeling with strong impression. Color chart is the basis of color designing. Using color charts can avoid the visual error which is caused by misunderstanding and intermingling colors. Furthermore, the appropriate colors can be determined by their color value and color hue, and selected up from the color charts. Unfortunately, most color charts are designed and represented in 2-dimension. Furthermore, it is very difficult for users to survey the whole color model and choose colors in an intuitive way. Color psychology is a science of chromatology and psychology. Different colors give people different feeling by the relation between color and psyche. It provides a practical and intuitive mapping function between colors and user feeling. In our system, we construct the user-customized color charts by performing the following procedures. First, users can define their current psychological emotion directly by using a 3D emotion map, or indirectly by choosing keywords such as the name of artist or painting from random tables. Then, related images are searched from the web by adopting a web mining technology. A appropriate colors are therefore extracted to construct the color chart by analyzing these images. Finally, the color chart is rendered in a 3D visualization system which allows users to view and manage the distribution of colors interactively.