摘要
超分辨率图像重建技术可以提高图像分辨率,但是通常会给图像带来相应的尺寸变化,如何评价质量提升是个难题。目前常用的图像质量评价算法很少涉及图像的尺寸变化。基于结构相似度(SSIM)和尺度空间理论(SIFT),提出了一种针对超分辨率重建图像的弱参考质量评价算法,算法将SSIM与SIFT融合,能够适用图像尺寸的变化。通过仿真和实验证明了该算法的有效性。实验结果表明,该算法能够很好地适应图像尺寸的变化,可以客观地评价超分辨重建图像质量的好坏。
Super-Resolution(SR)imaging can increase the image resolution in spite of the detector limitation, but also bring the change of image size. In this paper, a Reduced-Reference Image Quality Assessment(RRIQA)algorithm based on the Structural Similarity Image Metric(SSIM)and Scale Invariant Feature Transform(SIFT)is proposed. This algorithm is also applicable to more referenced images. Experimental comparisons demonstrate the effectiveness of the proposed method and it is applicable to the condition of size changing.
作者
于康龙
秦卫城
杨进
许若飞
YU Kanglong;QIN Weicheng;YANG Jin;XU Ruofei(Unit 93498 of PLA, China;Unit 93501 of PLA, China)
出处
《计算机工程与应用》
CSCD
北大核心
2017年第2期201-205,共5页
Computer Engineering and Applications
关键词
图像质量评价
超分辨图像重建
结构相似度
尺度空间理论
image quality assessment
imaging super-resolution reconstruction
Structural Similarity Image Metric(SSIM)
Scale Invariant Feature Transform(SIFT)
