A no-reference sharpness metric sensitive to blur and noise

Abstract

A no-reference objective sharpness metric detecting both blur and noise is proposed in this paper. This metric is based on the local gradients of the image and does not require any edge detection. Its value drops either when the test image becomes blurred or corrupted by random noise. It can be thought of as an indicator of the signal to noise ratio of the image. Experiments using synthetic, natural, and compressed images are presented to demonstrate the effectiveness and robustness of this metric. Its statistical properties are also provided.

Evaluation of image quality metrics for sharpness enhancement

“…2) No reference metrics: Recent NR metrics are based on the assumption that the distortion types are known (such as blocking artifacts [28], blur and noise [29], [30], [31], JPEG [32] or JPEG2000 compression [33], [34], and others [35], [36]). There is an important notion proposed by Ferzli and Karam: Just-Noticeable Blur (JNB) [37], [38], which takes into account the response of the HVS to sharpness at different contrast levels.…”

Restoration for weakly blurred and strongly noisy images

“…An image sharpness metric, which can estimate the local image blur even in the presence of noise [20,19], is implemented here. The metric is based on the singular values of local gradient matrix, which have already been calculated in the SK construction stage, so it does not require much extra calculation.…”

“…The metric is based on the singular values of local gradient matrix, which have already been calculated in the SK construction stage, so it does not require much extra calculation. The local metric Q for the pixel located at x l is defined as [20,19]:…”

Learning a blind image quality index based on visual saliency guided sampling and Gabor filtering

“…DS approaches assume that the distortion type is known as a prior knowledge. These methods mainly measure the impact of one distortion type on the image quality such as blur [1][2] or ringing [3]. One important feature of DS approaches, which limits their application domain, is that they are distortion-specific.…”