Multi-scale Structural Similarity Index Research Articles

An Optical Flow-Based Full Reference Video Quality Assessment Algorithm.

We present a simple yet effective optical flow-based full-reference video quality assessment (FR-VQA) algorithm for assessing the perceptual quality of natural videos. Our algorithm is based on the premise that local optical flow statistics are affected by distortions and the deviation from pristine flow statistics is proportional to the amount of distortion. We characterize the local flow statistics using the mean, the standard deviation, the coefficient of variation (CV), and the minimum eigenvalue ( λ min ) of the local flow patches. Temporal distortion is estimated as the change in the CV of the distorted flow with respect to the reference flow, and the correlation between λ min of the reference and of the distorted patches. We rely on the robust multi-scale structural similarity index for spatial quality estimation. The computed temporal and spatial distortions, thus, are then pooled using a perceptually motivated heuristic to generate a spatio-temporal quality score. The proposed method is shown to be competitive with the state-of-the-art when evaluated on the LIVE SD database, the EPFL Polimi SD database, and the LIVE Mobile HD database. The distortions considered in these databases include those due to compression, packet-loss, wireless channel errors, and rate-adaptation. Our algorithm is flexible enough to allow for any robust FR spatial distortion metric for spatial distortion estimation. In addition, the proposed method is not only parameter-free but also independent of the choice of the optical flow algorithm. Finally, we show that the replacement of the optical flow vectors in our proposed method with the much coarser block motion vectors also results in an acceptable FR-VQA algorithm. Our algorithm is called the flow similarity index.

Multi-scale structural similarity index for motion detection

The most recent approach for measuring the image quality is the structural similarity index (SSI). This paper presents a novel algorithm based on the multi-scale structural similarity index for motion detection (MS-SSIM) in videos. The MS-SSIM approach is based on modeling of image luminance, contrast and structure at multiple scales. The MS-SSIM has resulted in much better performance than the single scale SSI approach but at the cost of relatively lower processing speed. The major advantages of the presented algorithm are both: the higher detection accuracy and the quasi real-time processing speed.

Multiscale image quality measures for defect detection in thin films

Manufacturing defects in flat surface products such as thin films, paper, foils, aluminum plates, steel slabs, fabrics, and glass sheets result in degradation of the visual quality of the product image. This leads to less satisfied customers, waste of material, and bad company reputation. This research presents a novel application of image visual quality measures such as the multiscale structural similarity index (MS-SSIM). A novel algorithm has been implemented for fast detection and location of defects in many flat surface products. Comparison of the proposed algorithm with the state-of-the-art approaches indicate promising results. A defect detection accuracy of 99.1 % has been achieved with 98.62 % precision, 97.7 % recall/sensitivity, and 100 % specificity. The discriminant power shows how well the MS-SSIM discriminates very effectively between normal and abnormal surfaces. The MS-SSIM has resulted in much better performance than the single-scale SSI approach but at the cost of relatively lower processing speed. The major advantages of the presented approach are as follows: scale invariance, avoiding the problem of parameter selection in the case of the state-of-the-art Gabor filter banks based approach, the higher detection accuracy, and the quasi real-time processing speed.

Virtual single-source computed tomography using dual-source acquisition: a new technique for the dose-neutral intraindividual comparison of different scan protocols.

The objective of this study was to compare the image quality of a standard single-source (SSS) computed tomography (CT) with that of a virtual single-source CT (VSS-CT) data set reconstructed from 2 raw data sets obtained by dual-source CT acquisition in abdominal CT to establish a radiation dose-neutral approach for the intraindividual comparison of 3 acquisition protocols at different radiation dose levels (RDLs). An abdominal phantom representing an 80-kg male was imaged using dual-source CT (SOMATOM Definition; Siemens Healthcare) at 3 RDLs with 120 kV(p) and different tube currents (low, standard, and high milliampere-second protocol). For each RDL, raw data were obtained once in single-source mode using x-ray tube A only and 5 times in dual-source mode using different ratios for tube current of x-ray tubes A and B (same total radiation dose; A/B: 90%/10%, 80%/20%, 70%/30%, 60%/40%, 50%/50%). For each RDL, SSS-CT and 5 virtual single-source image data sets (VSS-CT50 - 90) were reconstructed. To compare SSS-CT and VSS-CT data sets, image quality was assessed in terms of high- and low-contrast performance by calculating the modulation transfer function, image noise, noise power spectrum, and, for low contrast lesion detectability, the modified multiscale structural similarity index (MS-SSIM*). A maximum decrease of Δ = 5% of image quality compared with SSS-CT was defined as acceptable, and a noninferiority analysis with Δ was performed. For modulation transfer function, noninferiority was observed for all VSS-CT data sets and RDL (P < 0.05). Image noise demonstrated an acceptable increase (<3.2%, P < 0.05) for each RDL and noise power spectrum showed only minor differences in the midfrequency range. The MS-SSIM* index demonstrated for the high RDL protocol a minor decrease for VSS-CT data sets (<2%, P < 0.05). For the standard and low RDL, the relative differences of the MS-SSIM* index increased and were only in 1 case above Δ (standard RDL, mean VSS-CT80 5.1%, P > 0.05). The image quality obtained by virtual and SSS reconstruction using equivalent total radiation exposure to the patient showed only negligible differences in image quality. Therefore, this technique might allow an intraindividual comparison of full and reduced radiation dose protocols within 1 image acquisition step by simply splitting the radiation dose between the 2 x-ray tubes of a dual-source CT.

Salt and Pepper Noise Detection and Removal in Gray Scale Images: An Experimental Analysis

Impulse noise removal is a mechanism for detection and removal of impulse noise from images. Median filters are preferred for removing impulse noise because of their simplicity and less computational complexity. In this paper, impulse noise removal using the standard median filter and its variants are analyzed. Extensive simulations have been carried out on a set of standard gray scale images and the state of the art median filter variants are compared in terms of the well known image quality assessment metrics namely mean square error, peak signal to noise ratio and multiscale structural similarity index.

Color image quality assessment algorithm based on color structural similarity

Concerning the disadvantages of quality assessment algorithms for color images,a new algorithm based on visual structural similarity was proposed.Firstly,testing images were transformed into a selected uniform color space LAB2000HL.Secondly,one luminance Contrast Sensitivity Function(CSF) and two chromatic CSFs were used to filter images respectively.Thirdly,three structural similarity indexes were computed by multi-scale structural similarity index measurement(M-SSIM).Lastly,the proposed algorithm was constructed by weighting the three structural similarity indexes,which depended on different visual sensitivities of luminance and chromaticity.In the experiment,testing results on TID2008 database were compared with the results of visual assessment by Spearman rank-order correlation coefficient and Kendall rank-order correlation coefficient.The experimental results show that the proposed algorithm is more consistent with visual assessment and outperforms several other popular image quality assessment algorithms.

Influence of Sinogram Affirmed Iterative Reconstruction of CT Data on Image Noise Characteristics and Low-Contrast Detectability: An Objective Approach

ObjectivesTo utilize a novel objective approach combining a software phantom and an image quality metric to systematically evaluate the influence of sinogram affirmed iterative reconstruction (SAFIRE) of multidetector computed tomography (MDCT) data on image noise characteristics and low-contrast detectability (LCD).Materials and MethodsA low-contrast and a high-contrast phantom were examined on a 128-slice scanner at different dose levels. The datasets were reconstructed using filtered back projection (FBP) and SAFIRE and virtual low-contrast lesions (-20HU) were inserted. LCD was evaluated using the multiscale structural similarity index (MS-SIM*). Image noise texture and spatial resolution were objectively evaluated.ResultsThe use of SAFIRE led to an improvement of LCD for all dose levels and lesions sizes. The relative improvement of LCD was inversely related to the dose level, declining from 208%(±37%), 259%(±30%) and 309%(±35%) at 25mAs to 106%(±6%), 119%(±9%) and 123%(±8%) at 200mAs for SAFIRE filter strengths of 1, 3 and 5 (p<0.05). SAFIRE reached at least the LCD of FBP at a relative dose of 50%. There was no statistically significant difference in spatial resolution. The use of SAFIRE led to coarser image noise granularity.ConclusionA novel objective approach combining a software phantom and the MS-SSIM* image quality metric was used to analyze the detectability of virtual low-contrast lesions against the background of image noise as created using SAFIRE in comparison to filtered back-projection. We found, that image noise characteristics using SAFIRE at 50% dose were comparable to the use of FBP at 100% dose with respect to lesion detectability. The unfamiliar imaging appearance of iteratively reconstructed datasets may in part be explained by a different, coarser noise characteristic as demonstrated by a granulometric analysis.

Full-Reference Image Quality Metrics Performance Evaluation Over Image Quality Databases

A quantitative predictive performance evaluation of 18 well-known and commonly used full-reference image quality assessment metrics has been conducted in the present work. The process has been run over six publicly available and subjectively rated image quality databases for four degradation types namely JPEG and JPEG2000 compression, noise and Gaussian blur. Results show that the existing predictive performance evaluation tools of the different full-reference image quality metrics are significantly impacted by the choice of the image quality database. Three of them, namely Toyama, LIVE and TID, have been found to give different assessment results. The visual information fidelity (VIF) quality metric has been found to have superior predictive capabilities to its counterparts. MS-SSIM (multi-scale structural similarity index), MSSIM (modified SSIM) and VIFP (pixel-based VIF) have also closer performances in terms of their correlation to the subjective human ratings, accuracy and monotonicity to the VIF model.

Interference Shaping for Improved Quality of Experience for Real-Time Video Streaming

The unpredictability of the wireless medium poses a major challenge to delivering a high quality of experience (QoE) for real-time video services. Bursty co-channel interference is a prominent cause of wireless throughput variability, which leads to video QoE degradation, even for a fixed average channel quality. In this paper, we propose and analyze a network-level resource management algorithm termed interference shaping to smooth out the throughput variations (and hence improve the QoE) of video users by decreasing the peak rate of co-channel best effort users. Wireless link capacity variations are mapped to the real-time video packet loss rate, and the interference shaping QoE gain for video users is quantified by benchmarking against a modified multi-scale structural similarity (H-MS-SSIM) index. H-MS-SSIM is an accurate perceptual video quality metric that incorporates the important hysteresis effect whereby the current QoE (which is subjective) may strongly depend on the recent past. The proposed technique increases mean QoE and reduces the QoE variability over time, with a net perceptual increase of about 2-3x in illustrative settings while incurring insignificant decrease in the QoE for co-channel best effort users. Interference shaping can be implemented in both unicast and multicast real-time video streaming with much higher potential gains for multicast.

A systematic approach towards the objective evaluation of low-contrast performance in MDCT: Combination of a full-reference image fidelity metric and a software phantom

ObjectivesTo assess the feasibility of an objective approach for the evaluation of low-contrast detectability in multidetector computed-tomography (MDCT) by combining a virtual phantom containing simulated lesions with an image quality metric. Materials and methodsA low-contrast phantom containing hypodense spheric lesions (−20HU) was scanned on a 64-slice MDCT scanner at 4 different dose levels (25, 50, 100, 200mAs). In addition, virtual round hypodense low-contrast lesions (20HU object contrast) based on real CT data were inserted into the lesion-free section of the datasets. The sliding-thin-slab algorithm was applied to the image data with an increasing slice-thickness from 1 to 15 slices. For each dataset containing simulated lesions a lesion-free counterpart was reconstructed and post-processed in the same manner. The low-contrast performance of all datasets containing virtual lesions was determined using a full-reference image quality metric (modified multiscale structural similarity index, MS-SSIM*). The results were validated against a reader-study of the real lesions. ResultsFor all dose levels and lesion sizes there was no statistically significant difference between the low-contrast performance as determined by the image quality metric when compared to the reader study (p<0.05). The intraclass correlation coefficient was 0.72, 0.82, 0.90 and 0.84 for lesion diameters of 4mm, 5mm, 8mm and 10mm, respectively. The use of the sliding-thin-slab algorithm improves lesion detectability by a factor ranging from 1.15 to 2.69 when compared with the original axial slice (0.625mm). ConclusionThe combination of a virtual phantom and a full-reference image quality metric enables a systematic, automated and objective evaluation of low-contrast detectability in MDCT datasets and correlates well with the judgment of human readers.

Structural Similarity Measure for Color Images

images reveal more meaningful information to the human observers rather than grayscale ones. Regardless of the advantages of the existing well-known objective image quality measures, one of the common and major limitations of these measures is that they evaluate the quality of grayscale images only and don't make use of color information. In this paper we propose an improved method for image quality assessment that adds a color comparison to the criteria of the well-known Multiscale Structural Similarity index (MSSIM). We evaluated the new color image quality measure through human subjective experiments. Our human subjective evaluation data contains 25 reference images and 875 test images produced by five popular color quantization algorithms. Each of the quantized images was evaluated by twenty two subjects and more than 19200 individual human quality judgments were carried out to obtain the final mean opinion scores. We also tested the proposed method on TID2008 image database to further verify our results. These results indicate that adding color comparison improves MSSIM for many distortions in TID2008 and for assessing quantized images in our database. Keywordsquality assessment, Structural similarity index, Color quantization.

Directional lifting wavelet and universal trellis-coded quantisation-based image coding algorithm and objective quality evaluation

In this study, an image coding algorithm based on directional lifting wavelet transform (DLWT) and universal trellis-coded quantisation (UTCQ) is presented, and the coding performance is evaluated with three objective image quality metrics. Compared to the discrete wavelet transform, DLWT performing prediction and update along the direction of the local region can provide an efficient representation of edges in images, but shows a similar ability in representing the smooth region. To further improve the visual quality of the smooth background regions, UTCQ is adopted to quantising the wavelet coefficients. The proposed algorithm is measured with not only the dominant peak signal-to-noise ratio (PSNR), but also new metrics multi-scale structural similarity index measure (MSSIM) and visual information fidelity (VIF) which provide a better approximation to the perceived image quality than PSNR by taking the property of human visual system (HVS) into account. Experimental results show that the proposed algorithm has the best MSSIM and VIF performance among the compared algorithms (including JPEG2000) for the typical test images, and its decoded images at low bit-rate are visually more appealing in both edges and smooth background regions. For image Barbara, the proposed algorithm outperforms JPEG2000 up to 24.63% relatively in VIF and 1.93%dB in PSNR at 0.5%bpp, at most 3.62% relatively in MSSIM at 0.125%bpp. The experimental results also show that UTCQ does perform better than scalar quantisation (SQ) in MSSIM and VIF and improves the subjective visual quality, although UTCQ is not necessarily better than SQ in PSNR.