數位向量分析應用於新型顯示器之色彩特性化研究

Abstract

隨著近年顯示器技術的快速發展，人們對於顯示器的要求也越來越多元。輕、薄、高畫質、高對比、色彩豔麗等已經是基本的訴求，超低耗電、陽光下保有閱讀性、可撓曲等新興議題更是廣泛的被大家討論。研究者與製造商無所不用其極的設計出各式各樣精巧複雜的顯示系統，只為了滿足大家的需求，電子紙便是一個很好的例子。電子紙因為具有雙穩態、超低耗能、強光下維持對比、類似紙張的閱讀感受、容易大量製造以及具有可撓性的特點，近年備受期待能成為次世代顯示器應用，如電子閱讀器、電子標籤、電子報和電子看板等等。 新型顯示器身為消費者終端的閱讀媒介，建立一套精確的色彩特性化模型極為重要，如此才能與其他傳統顯示設備進行良好的色彩重建。然而過去所習之色彩特性化方法，如受限於複雜的色料混合；如數值模型及對照表模式僅針對兩個色彩空間映射而不具有任何物理意義，並不能有效的針對多樣化顯色機制的新型顯示器建立特性化。 因此本論文提出一種數位向量分析的技巧，根據主成分分析法和獨立成分分析法，將所量測到新型顯示器的頻譜資訊進行資料處理、基底轉換、特徵值拆解、設定評價函數、遞迴優化等，即可找到於新型顯示器當中最主要的色料頻譜成分或是獨立的色料頻譜成分，並以此成分為系統的轉換矩陣，建立一套屬於新型顯示器的色彩特性化。本論文最終以此技術分別應用於單色與全彩的電泳式顯示器，成功的分析於此複雜系統當中的特徵頻譜為何，並完成一能精確描述此電子紙顯色之色彩模型。

Technologies in displays have made a rapid progress in the last decade. At this time, great technological advances in displays are profoundly changing the way image is generated. These promising novel displays are capable of fulfilling diverse requirements individually such as lightness, thinness, high image quality, 3D image, energy saving, fine contrast under sunlight, flexibility and so on. To be an output device for end user, it should be the most vital issue that having a complete colorimetric characterization will benefit a precise color reproduction across all kinds of novel displays. However, conventional techniques cannot be sufficiently applied into all of novel display due to the complicated rendering mechanism. In view of this, vector component analysis is introduced to solve this problem. In this doctoral research, we mainly focus on vector component analysis − in particular, the methodologies based on principal component analysis and independent component analysis. Through preprocessing data, change of basis, eigenvalue decomposition, defining merit function, and optimizing with iteration, the principal or independent components representing the specifically spectral features of novel display can be extracted from the measured spectral information. In addition, the proposed analyzing techniques will be further applied into three cases: monochrome micro-cup electrophoretic display, full-color micro-cup electrophoretic display, and full-color micro-encapsulated electrophoretic display. According to the experimental results, the featuring spectra can be successfully determined by PCA-based and ICA-based methodologies. Consequently, a colorimetric characterization of the special EPD with high fidelity has been done.