No-reference image quality assessment is a challenging task due to the absence of a reference image in practical situations to quantify image quality. This paper proposes a new no-reference image quality metric for natural images using latent noise estimation, Gabor response, and contrast deviation. The algorithm employs an extension of gradient-based SSIM into the no-reference application using SVD-based AWGN estimation, and defines attributes such as Gabor-based smoothness and contrast deviation. The proposed metric arrives at an overall quality score by computing a linear weighted summation of the three image attributes. The proposed algorithm has been tested on several public databases (i.e. LIVE, TID 2013 and CSIQ), and the overall results display a noteworthy correlation of nearly 80% with the human visual system.

No-reference image quality assessment using Gabor-based smoothness and latent noise estimation

Distinguishing artwork from digital photographs is a simple task for a human observer. However, imparting this ability to a machine requires quantification of the quality of naturalness. In particular, the quality of ‘glossiness’ of a scene in a digital photograph as opposed to that in artwork or a painting of the same scene, is a challenging parameter to quantify. In this paper, a classification-based approach is proposed to quantify this glossiness of an image. The authors propose and validate the hypothesis that features extracted from the curvelet transform contain information regarding glossiness of an image. By training a support vector machine (SVM) classifier using curvelet coefficient features of images of varying naturalness, a framework to predict no-reference glossiness index for any general query image is proposed. The reliability of the proposed metric was then gauged by obtaining its correlation with subjective scores on naturalness provided by human observers. Results exhibit noteworthy performance of this metric in representing image glossiness, and could be further improved by using a larger training database and advanced classifiers.

Image glossiness from curvelet features using SVM-based classification

Noise reduction has an impact on preprocessing and quality enhancement of images. In this paper noise reduction techniques are applied on Estampages. Estampages are the exact replica of inscriptions and are the source of our ancient history. In preserving and processing for future, image processing techniques are applied. Noise reduction is mainly accomplished using filters, in our study various filters like Median, Gaussian, Wiener, Gabor, Box filtering, order statistics, morphological operations are applied and comparative study is made with respect to Structural similarity index. Suitable filters are combined with morphological operations to draw inference. Hybrid systems for noise reduction having Erosion combined with Wiener/Median followed by dilation processes are applied to test Estampage images. Visible and comparative studies using SSIM have evidenced good result for Erosion-Median-Dilation & Erosion-Wiener-Dilation combinations.

Comparative study of basic and hybrid filters for the reduction of noise in Estampages

Image quality assessment (IQA) plays a crucial role in monitoring quality control in image communication systems, and in benchmarking and optimizing parameters in enhancement algorithms. The full-reference IQA metrics require a good-quality reference image, obtaining which may not be practical in real-life applications. This paper, therefore, proposes a no-reference IQA metric based on the hypothesis that every image has latent additive white Gaussian noise (AWGN). A mathematical model was developed on a dataset of fifty test images by computing gradient-based structural similarity of corrupted images w.r.t. the original. Statistical modeling of the observations were found to fit an exponential parametric model. The standard deviation of the latent (or apparent) AWGN present in any image was estimated using an SVD-based approach. The proposed metric, referred to as the no-reference gradient-based structural integrity (NRGSI), is then computed by a simple backprojection of the estimated noise deviation on the exponential model. The accuracy of the proposed objective metric is characterized by its comparison with subjective quality scores given by ten subjects, and with a classical perceptual quality measure.

Vineet Kumar

Papers

No-reference image quality assessment using Gabor-based smoothness and latent noise estimation

No-Reference image quality assessment using gradient-based structural integrity and latent noise estimation