彩色電子紙之影片播放半色調演算法設計

Abstract

現有的半色調演算法在處理連續畫面時面臨到高圖像品質無法避免閃爍的問題。本研究提出一種基於直接二元搜尋演算法(Direct Binary Search, DBS)的可行解決方案，並以此進行了三項改進。首先，在演算法的執行效率上，通過空簇演算法(void-and-cluster)產生的偽隨機順序矩陣重新決定了DBS 演算法中像素的執行次序，大幅的減少了處理每一幀所需的時間，使其適用於實時的影片處理。其次，本研究亦改善了演算法處理結果的藍雜訊特性，通過隨機亂數重新確定每個順序矩陣的起始位置，進而讓新提出的方法獲得更好的影像品質，並使用徑向平均功率譜密度來驗證影像品質的改進。最後，本研究針對電子紙上的彩色濾光片排列來改良演算法的搜尋方式，不僅提高了演算法的搜尋效率，也顯著提高了處理後影像在電子紙上顯示時的色彩飽和度。

The halftoning algorithm faces the challenge of retaining image quality without flicker when processing image sequences. This study presents a feasible solution based on direct binary search algorithm with three improvements. First, in terms of the processing efficiency of the algorithm, the pixel execution order of DBS algorithm is re-determined through the pseudo-random class matrix generated by void-and-cluster algorithm, which tremendously reduces the time required to process each frame for the algorithm, making it suitable for real-time video processing. Second, this study also improves the blue noise property of the results, the starting position of each class matrix is re-determined through random numbers so that the proposed method achieves better image quality. Radially averaged power spectral density is also used to verify the improvements in image quality. Finally, this study modifies the algorithm's search method for the color filter array arrangement on electronic paper, which not only improves the search efficiency of the algorithm but also significantly improves the color saturation of the processed image.