單色影像反轉半色調法之設計

Abstract

反轉半色調法之目的在將半色調法產生的二元影像還原回灰階影像。一般 說來，半色調法是印刷界印製照片時常採用的步驟。然而，許多影像處理 方式如濾波、加強及壓縮，對於灰階影像的成效較佳。同時，反轉半色調 法可應用於監視印刷成品，有助於自動調整印刷程序。這使得反轉半色調 法在實際應用上益形重要。在本論文中，我們先提出兩種基本的反轉半色 調法和適當的後級處理法則。利用反系統模式來尋找具有最小誤差平方的 反轉半色調影像；再利用對影像特徵的了解，我們設計出的後級處理法能 更進一步的增進影像品質，使其更適於人類視覺。在觀念上，這和傳統上 僅用簡單低頻濾波的反轉半色調法有很大的差別。而後，我們發展出強韌 地適應性濾波器法，簡稱 SV-SWF。它的目標在使得基本反轉半色調法可 藉由適應參數調整而運用於不同型式的照片。這個方法是利用適當的影像 分類法將影像資料區分為彼此性質相異的集合，再根據事先訓練所得不同 的對應反轉濾波器參數加以還原之。經過詳細的分析評估顯示：無論在客 觀數值或主觀評價上，根據以上方法所產生的還原影像品質均較已知的其 他反轉半色調法為佳。 Two basic inverse halftoning methods together with some post- processing schemes are proposed in this thesis. In order to optimize the inverse operation of halftoning, we first take the inverse modeling concepts to find the least-squares solutions of this problem. And then, the knowledge of image characteristics can help us in designing improved algorithms that produce the most promising pictures to our eyes. They are conceptually very different from the traditional inverse half- toning methods that use simple low-pass filters. Furthermore, we develop a general space-varying sliding-window filter (SV- SWF) scheme. The goal is to handle various types of pictures using the same basic inverse halftoning structure together with adaptive parameters. This universal inverse halftoning operator is achieved by using an adequate classification scheme to separate image data into different groups, each corresponding to a set of pre-trained parameters. It is proven through detailed analysis on the performance of the proposed algorithms that these algorithms produce better reconstructed images than the other known algorithms in both PSNR and subjective evaluation.