- 期刊
影像增強於跨媒體色彩管理系統之應用
The Application of Image Enhancement on Cross-Media Color Management System
摘要
一般影像處理中所用到的影像增強模式，僅針對影像所在的色彩空間作直接的增強（如RGB或CMYK）。如此不但未考慮到影像在跨媒體複製時所會遇到的問題，也可能使增強的影像之色相有所改變。本研究中結合多媒體色彩管理系統並加入對於影像增強之研究，達到設備獨立、光源獨立的影像增強演算法。研究中對四種不同的影像增強模式（對比延伸法、S曲線對比增強、銳利化、反銳化遮罩）經由心理物理實驗以及S-CIELAB色差公式作效能比較，找出最佳的影像增強模式。 我們將影像所在的設備從屬色彩空間搭配使用CIECAM02色外貌模式產生的ICC profile轉換到設備獨立的色外貌CIEJCh色彩空間，再對影像的明度屬性及彩度屬性作影像增強的轉換，如此可以達到跨媒體複製的要求，也能避免增強後影像色相改變的問題。 結果顯示反銳化遮罩(Unsharp Mask)具有最佳的效能表現，表現比未經處理的影像更好。
並列摘要
Traditionally, in the field of image processing application, models, used in image enhancement, only enhance the images in interest in its device dependent color space (e.g. RGB or CMYK). Hence, the reproduced results on the enhanced images, obtained via cross-media reproduction system, often have hue changes. In this research, by the combination of image enhancement, totally four algorithms were derived based on both device independence and illuminant independence. Those were algorithms of contrast stretching, s-curve, sharpness, and unsharp mask. Also a set of psychophysical experiments and S-CIELAB was conducted to evaluate performances of models derived, and to find the best-performing image enhancement model. Firstly, images tested were transferred from the image's device dependent color space to device independent color space CIEJCh using their corresponding ICC profile which had been produced by using CIECAM02 color appearance model. And then, in terms of luminance and chroma factor, images produced were further enhanced via algorithms of image enhancement algorithms derived in this study. Therefore, not only the request of crossmedia reproduction had been reached, but also the changes of hue in enhanced image had been avoided. Totally, the image enhancement model of unsharp mask outperformed the others in this study, and gave better rendition-quality of images than those which had not been enhanced.