No-reference perceptual blur model based on inherent sharpness

摘要: An objective blurriness model is useful in various image processing applications. Especially, a no-reference expected as highly desirable approach due to its applicability wide range of Blurriness an known be commonly induced by the attenuation spatial high-frequency and thus most conventional researches focused on estimating amount high-frequency. However, human-perceived might varied across contents. Very few have been investigated human visual system for perception. To address lack efficient model, this paper presents perception designed spatially varying function based inherent sharpness. Pixel-wise perceptual computed employing then integrated into overall index using saliency information. The experimental comparisons with state-of-the-arts models extensive public databases show that proposed well-correlated subjective scores different content images, outperforms compared models.