基於線段扭曲與幾何變形之影像濃縮畫質衡量機制;Quality Assessment of Image Retargeting based on Line Bending and Geometric Distortion

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/74697

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/74697

Title:	基於線段扭曲與幾何變形之影像濃縮畫質衡量機制;Quality Assessment of Image Retargeting based on Line Bending and Geometric Distortion
Authors:	方志筠;Fang, Chih-Yun
Contributors:	資訊工程學系
Keywords:	影像畫質衡量;畫面扭曲;重要線段;SIFT Flow;Image quality assessment;Distortion;Important line;SIFT Flow
Date:	2017-07-27
Issue Date:	2017-10-27 14:36:35 (UTC+8)
Publisher:	國立中央大學
Abstract:	影像濃縮(retargeting)的目標是讓具有固定尺寸的影像資料在各種解析度的畫面輸出中都能有良好的成像。然而，不同的影像濃縮方法往往對於調整後影像有相異的效果，因此接近使用者主觀感受的畫質衡量機制是必要的。本研究針對在具有原始影像做為參考下的濃縮影像提出客觀的畫質衡量機制，本機制分成兩個部分，包括重要線段扭曲(Line Bending)與幾何形狀變形(Geometric Distortion)，兩者都會利用SIFT Flow 去判別像素點在兩張圖中的的位移情形。我們主要檢視SIFT Flow 的變化情況決定影像的幾何變形程度，另一方面根據視覺顯著圖(Saliency Map)在畫面上標示重要線段，再以SIFT Flow 找出線段扭曲位置，當幾何變形被判定失效時改以線段扭曲程度評量畫質。評量的參數設定儘量貼近平均意見分數(MOS)，期望與人眼的主觀感受一致。實驗結果測試了不同影像與濃縮方法的評估結果，並與其他的客觀衡量方式比較，以顯示我們所提出的機制符合主觀感受的準確程度。;Image retargeting is a technique to output images with a different aspect ratio from those of displaying devices. Various methods exist but they may not be consistent with different kinds of images. It is essential to develop good quality assessment approaches in this field. In this research, we propose an objective quality assessment for image retargeting with original images as the reference. The proposed scheme includes two parts: line bending and geometric distortion, both of which are based on SIFT Flow to examine the degree of pixel shifting. The proposed scheme basically employs the variation of SIFT Flow to determine the geometric distortion. At the same time, some important lines are drawn according to saliency map to determine line distortion, which is used to grade the quality when the geometric distortion is not reliable. The setting of the parameters aims at making the score closer to the mean opinion scores (MOS), which represents that the quality assessment is consonant with human’s vision. Experimental results show the accuracy of the proposed scheme by comparing with other objective image quality assessment methods and different retargeting approaches.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	219	View/Open

社群 sharing

Loading...