影像品質最佳化之研究 -以影像銳利度調整為例

CCUR > College of Journalism and Mass Communications > Department of Information Communications & Graduate Institute of Information Communications > thesis > Item 987654321/19919

Please use this identifier to cite or link to this item: https://irlib.pccu.edu.tw/handle/987654321/19919

Title:	影像品質最佳化之研究 -以影像銳利度調整為例
Authors:	歐子平
Contributors:	資訊傳播學系
Keywords:	影像品質影像銳利度掃描機銳利度配對比較法複迴歸分析
Date:	2010
Issue Date:	2011-10-27 12:25:41 (UTC+8)
Abstract:	隨著科技快速發展，人們對於影像品質的要求也越來越高。更佳影像品質的探究一直都是影像處理領域的研究重點。影響影像品質的因素相當多，最主要的三個因素分別為階調、色彩與銳利度，其中銳利度直接影響到視覺對影像優劣之判斷，對於影像品質具有相當重要的影響，傳統上對銳利度的調整只著重影像後製處理。本研究以探討影像銳利度對影像品質的影響為主，觀察影像擷取端的銳利度與Photoshop遮色片銳利化調整之間的相關，使用掃描機當作影像擷取端，掃描彩色正片為實驗影像。透過主觀視覺評估方法來找出較符合視覺偏好的影像擷取與影像後製之間的銳利度最適化控制與迴歸方程式。本研究使用Thurstone配對比較法進行心理物理的主觀視覺評估，與客觀影像評估指標進行統計分析。實驗結果得知，影像在擷取端與後製都需進行適當的銳利化處理，而影像銳利度視覺偏好結果呈現倒U曲線的關係，表示可定出最佳化的銳利度調整區間。本研究也計算出銳利度調整的複迴歸方程式，與客觀影像評估對視覺銳利度的迴歸方程式，提供後續影像銳利度調整使用時的參考依據。 The demand for better image quality becomes higher and higher, and has been the focal point in the field of image processing with the rapid development in imaging technology. Many factors affects the image quality. Three of the most important factors are “Color”, “Tone” and “Sharpness”. “Image Sharpness” would effect the visual perception on image quality directly. It is a very important factor to be considered when optimizing the image quality. Traditionally, Sharpness adjustment is only emphasized in the image post-processing stage. This study aims to investigate how the sharpness affected the image quality by using a high-end flatbed scanner to scan color positive film and then adjusting the image sharpness both in the scanner setting as well as in Photoshop Unsharp Mask setting, and identifying the image quality by visual assessment. The pair-comparison method derived using Thurstone’s Law of Comparative Judgment is used in this study as visual assessment to establish interval scale for image quality index. The image sharpness visual preferences showed an “Inverted U“ shape, which means the most appropriate setting can be found on the peak. The experimental results indicated that the image needed to be sharpened both in the scanning and the post-processing stages. Regression equations were also derived between the image sharpness adjustment and the visual assessment as a reference to adjust the image sharpness.
Appears in Collections:	[Department of Information Communications & Graduate Institute of Information Communications ] thesis

Files in This Item:

File	Description	Size	Format
gsweb357.pdf		9694Kb	Adobe PDF	2697	View/Open

Loading...