TY - GEN
T1 - An image quality assessment algorithm based on feature selection
AU - Lu, Ting
AU - Zhang, Yanning
AU - Li, Haisen
PY - 2013
Y1 - 2013
N2 - Image Quality Assessment(IQA) is of fundamental importance to numerous imaging and video processing applications. For most of the applications, the perceptual meaningful measure is the one which can automatically assess the quality of images or videos in a perceptually consistent manner. However, most commonly used IQA metrics are not consistent well with the human judgments of image quality. Recently, the SSIM metric which takes people's visual characteristics into consideration performs much better than the traditional PSNR/MSE. But the defects of it still exit on some specific kinds of distortions. A new algorithm of IQA based on feature selection is proposed in this paper. Local gradient entropy and phase congruency are added to the SSIM framework. Through in-depth feature selection and definition plus better pooling strategy, this algorithm performs much better in LIVE datasets.
AB - Image Quality Assessment(IQA) is of fundamental importance to numerous imaging and video processing applications. For most of the applications, the perceptual meaningful measure is the one which can automatically assess the quality of images or videos in a perceptually consistent manner. However, most commonly used IQA metrics are not consistent well with the human judgments of image quality. Recently, the SSIM metric which takes people's visual characteristics into consideration performs much better than the traditional PSNR/MSE. But the defects of it still exit on some specific kinds of distortions. A new algorithm of IQA based on feature selection is proposed in this paper. Local gradient entropy and phase congruency are added to the SSIM framework. Through in-depth feature selection and definition plus better pooling strategy, this algorithm performs much better in LIVE datasets.
KW - Feature selection
KW - Image quality assessment
KW - Local gradient entropy
KW - Phase congruency
UR - http://www.scopus.com/inward/record.url?scp=84892930628&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-36669-7_36
DO - 10.1007/978-3-642-36669-7_36
M3 - 会议稿件
AN - SCOPUS:84892930628
SN - 9783642366680
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 289
EP - 297
BT - Intelligent Science and Intelligent Data Engineering - Third Sino-Foreign-Interchange Workshop, IScIDE 2012, Revised Selected Papers
T2 - 3rd Sino-Foreign-Interchange Workshop on Intelligent Science and Intelligent Data Engineering, IScIDE 2012
Y2 - 15 October 2012 through 17 October 2012
ER -