Structural Toughness Under Noise: An Efficient No-Reference Image Distortion Assessment for Blur and Noise
- PDF / 2,546,557 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 3 Downloads / 186 Views
ORIGINAL ARTICLE
Structural Toughness Under Noise: An Efficient No‑Reference Image Distortion Assessment for Blur and Noise So‑Yeong Jeon1 · Daeyeon Kim1 Received: 31 December 2019 / Revised: 19 February 2020 / Accepted: 16 April 2020 / Published online: 18 May 2020 © The Korean Institute of Electrical Engineers 2020
Abstract In image denoising and reconstruction problems, it is useful to monotonically quantify the distortions in images representing the same scene. For this purpose, we propose a training-free no-reference image distortion assessment method for both blur and noise. The method is based on the observation that the structural similarity between an input image and its shifted-andnoised copy is related to the levels of blur and noise distorting the input image. Computing a noised copy would require a random number generation, but assuming virtual noise independent from the input image, we derived the method that does not require computing actual noised copy. In our experiments of assessing the singly/multiply distorted images representing the same scene, the proposed method generally showed better monotonicity than competing state-of-the-art methods. The proposed method runs about 80 times faster than those competing methods and it can assess local regions of the input image, which makes it useful in spatially adaptive denoising applications. Keywords No-reference image distortion assessment · Multiple distortions · Blur · Noise · Structural similarity
1 Introduction Quantifying distortion within a digital image is important for image acquisition, transmission, compression, restoration, enhancement, etc. There are many kinds of distortions: blur, noise, compression, intensity shift, chromatic aberration, etc. Among the various types of distortions, blur and noise are frequently observed. Blur occurs due to out-of-focus, motion, low resolution of the camera, etc. On the other hand, noise occurs due to insufficient light energy, non-uniformity of pixel responses, electrical interference, etc. Recent researches for quantifying image distortions are conducted in terms of objective image quality assessment (IQA), which is aimed for predicting the image quality that would be perceived by an average human. IQA methods are largely classified into full-reference IQA (RIQA) methods [1–5] and no reference IQA (NR IQA) methods [6–15, 24–26]. RIQA methods predict the quality of a distorted
* So‑Yeong Jeon [email protected] Daeyeon Kim [email protected] 1
Agency for Defense Development, Daejeon, Korea
image by comparing it with its reference, i.e., distortionfree image, whereas NR RIQA methods predict the quality without the reference image. a well-known RIQA method is structural similarity index (SSIM) [1]. SSIM is designed based on local mean, local variance, and local covariance of the distorted image and its reference image. Motivated by the success of SSIM, many other RIQA methods [2–5] followed the framework of SSIM. The idea of the proposed method is also inspired by SSIM. In practical applications, howe
Data Loading...
