In the real application, digital images may undergo the process of acquisition, compression, and transmission, which causes the excess blurring, quantization, and noise. However, the metrics of image quality assessment for multiply distorted images are very limited. In this paper, we propose a new multi-scale learning quality-aware features blind image quality assessment algorithm for multiply distorted images by using both local phase and local amplitude. In the new model, a distorted image is decomposed into three scales by Gabor transform, and its phase congruency image (PCI), phase congruency covariance maximum image (PCCmax), and phase congruency covariance minimum image (PCCmin) are produced. Then, we calculate contrast sensitivity function and gray level-gradient co-occurrence matrix features from distorted image and its PCI, PCCmax, and PCCmin, and mean value of intensity of PCI, PCCmax, PCCmin, and overlapping blocked local amplitude of distorted image. At last, SVR is used to build the approximating function between these features and subjective mean opinion scores. Both local phase and local amplitude features are extracted at multi-scale images, which supply more flexibility than the previous single-scale methods in incorporating the variations of viewing scene. Comparative experiments between our proposed metric and the state-of-the-art full-reference and no-reference metrics are conducted on two newly released multiply distorted image databases (LIVEMD and MDID2013) that demonstrate the effectiveness of our proposed method.

A Multi-Scale Learning Local Phase and Amplitude Blind Image Quality Assessment for Multiply Distorted Images

In the paper the issue of quality assessment of 3D printed surfaces using image analysis is considered with particular attention paid to Fourier analysis of image fragments resized to one-dimensional vectors. Due to the application if Fourier analysis the regularity of visible patterns related to the consecutive layers of the filament can be assessed, assuming the side view of the printed surface. In order to avoid the problems of uneven lighting, our experiments have been conducted for scanned images of several 3D printed flat samples. As some of them have been contaminated by forced distortions, it is possible to classify them into two groups depending on the presence and amount of them. Due to the application of Fourier analysis some encouraging experimental results have been obtained which can be useful also for online monitoring of 3D prints quality for the images captured by cameras.

Quality Assessment of 3D Printed Surfaces in Fourier Domain

Since solving the inversion equation of synthetic aperture image reconstruction is an ill-posed mathematical problem, in order to meet the needs of high-precision reconstruction of scene brightness temperature images during ground detection, the mechanism analysis and algorithm research of high-resolution brightness temperature reconstruction is carried out. In this paper, various resolution-enhanced inversion algorithms such as statistical inversion are analyzed, and the synthetic aperture response is classified. The direct-current (DC) component of the synthetic aperture response is processed by regularized G matrix inversion, and the alternating (AC) component of the synthetic aperture is processed by Fourier inversion, which reduces the bandwidth of the point spread function, solves the problem of sensitivity deterioration when the resolution is increased, and improves the image quality and result correlation, enabling high-resolution image reconstruction. The overall research is verified by numerical calculation of the standard mean error and root mean square error of synthetic aperture inversion brightness temperature image.

A High-resolution Image Reconstruction Technology Based on G-matrix and Fourier Transform

Since solving the inversion equation of synthetic aperture image reconstruction is an ill-posed mathematical problem, in order to meet the needs of high-precision reconstruction of scene brightness temperature images during ground detection, the mechanism analysis and algorithm research of high-resolution brightness temperature reconstruction is carried out. In this paper, various resolution-enhanced inversion algorithms such as statistical inversion are analyzed, and the synthetic aperture response is classified. The direct-current (DC) component of the synthetic aperture response is processed by regularized G matrix inversion, and the alternating (AC) component of the synthetic aperture is processed by Fourier inversion, which reduces the bandwidth of the point spread function, solves the problem of sensitivity deterioration when the resolution is increased, and improves the image quality and result correlation, enabling high-resolution image reconstruction. The overall research is verified by numerical calculation of the standard mean error and root mean square error of synthetic aperture inversion brightness temperature image. 

Texture is one of the most important image's features, which has an important role in some image-processing applications such as segmentation and classification. The main problem is to obtain how much textures are reminded from reference image in the distorted image. To solve this problem image quality, assessment (IQA) plays main role and compares the amount of difference of texture between both images. This paper proposed a novel method to assess fully reference (FR) texture quality assessment, which is more sensible and sensitive to pixel arrangements. The grayscale images used for our experiments are BRODATZ imagery. The experimental results show superiority of proposed method compared to some popular texture assessment methods.

Assessment of grayscale image in terms of texture quality using reference image

Reduced-Reference (RR) image quality measures aim to predict the perceptual quality of distorted image using only partial information about the original image. In this paper, an effective Reduced-Reference image quality assessment algorithm based on FFT transform and Edge Similarity is introduced. The main design principle of the proposed method is choice of the best blocks of Image. After dividing the source images into blocks of 16×16 pixels, calculating the FFT Transform for each block, the FFT Transform gives best blocks of image. Next, the important features blocks of the image were recognized by Edge and the same actions were done on the image of distortions and finally, the similarity of both images was calculated. The experimental results on LIVE and CSIQ databases show that our RR proposed metric correlates well with the subjective quality scores, also in comparison with commonly used full-reference metric and with a state-of-the-art reduced reference.

Zhila Azimzadeh

Papers

Reduced-Reference image quality assessment based on 2-D discrete FFT and Edge Similarity