安庆师范学院学报:自然科学版
安慶師範學院學報:自然科學版
안경사범학원학보:자연과학판
Journal of Anqing Teachers College(Natural Science Edition)
2012年
3期
31-33
,共3页
图像质量评价%分数阶微分%微分阶数%小波变换%感觉容量
圖像質量評價%分數階微分%微分階數%小波變換%感覺容量
도상질량평개%분수계미분%미분계수%소파변환%감각용량
image quality assessment%fractional differential%differential order%wavelet transform%sense capacity
基于感觉容量的图像质量评价算法(SC)计算简单、性能良好。它的图像细节用拉普拉斯变换系数表示,但拉普拉斯变换对噪声敏感,而分数阶微分可以提升信号中高频成分,且对噪声不敏感。文中提出了一种基于分数阶微分(FDSC)的图像质量评价方法,该方法将分数阶微分取代SC模型中的拉普拉斯变换。实验结果表明,FDSC模型与LIVE图库上差异主观评价分(DMOS)的线性相关性优于PSNR、SC和结构相似度算法。
基于感覺容量的圖像質量評價算法(SC)計算簡單、性能良好。它的圖像細節用拉普拉斯變換繫數錶示,但拉普拉斯變換對譟聲敏感,而分數階微分可以提升信號中高頻成分,且對譟聲不敏感。文中提齣瞭一種基于分數階微分(FDSC)的圖像質量評價方法,該方法將分數階微分取代SC模型中的拉普拉斯變換。實驗結果錶明,FDSC模型與LIVE圖庫上差異主觀評價分(DMOS)的線性相關性優于PSNR、SC和結構相似度算法。
기우감각용량적도상질량평개산법(SC)계산간단、성능량호。타적도상세절용랍보랍사변환계수표시,단랍보랍사변환대조성민감,이분수계미분가이제승신호중고빈성분,차대조성불민감。문중제출료일충기우분수계미분(FDSC)적도상질량평개방법,해방법장분수계미분취대SC모형중적랍보랍사변환。실험결과표명,FDSC모형여LIVE도고상차이주관평개분(DMOS)적선성상관성우우PSNR、SC화결구상사도산법。
Image quality assessment based on sense capacity is of simple and good performance. Detail of image is defined by Laplacian transformation coefficient which is sensitive to noise while fractional differentials can increase medium frequency and high frequency of signals and isn' t sensitive to noise. The thesis puts forward a new quality assessment algorithm based on fractional differ- ential. The new method is validated with subjeetive quality scores on LIVE database. Experimental results show that the linear corre- lation coefficient between subjective and objective scores of the new method is superior to the algorithms of PSNR,SC and SSIM.