Algorithms for video quality assessment (VQA) aim to estimate the qualities of videos in a manner that agrees with human judgments of quality. Modern VQA algorithms often estimate video quality by comparing localized space-time regions or groups of frames from the reference and distorted videos, using comparisons based on visual features, statistics, and/or perceptual models. We present a VQA algorithm that estimates quality via separate estimates of perceived degradation due to (1) spatial distortion and (2) joint spatial and temporal distortion. The first stage of the algorithm estimates perceived quality degradation due to spatial distortion; this stage operates by adaptively applying to groups of spatial video frames the two strategies from the most apparent distortion algorithm with an extension to account for temporal masking. The second stage of the algorithm estimates perceived quality degradation due to joint spatial and temporal distortion; this stage operates by measuring the dissimilarity between the reference and distorted videos represented in terms of two-dimensional spatiotemporal slices. Finally, the estimates obtained from the two stages are combined to yield an overall estimate of perceived quality degradation. Testing on various video-quality databases demonstrates that our algorithm performs well in predicting video quality and is competitive with current state-of-the-art VQA algorithms.
In this paper, we present a no-reference quality assessment algorithm for JPEG2000-compressed images called
EDIQ (EDge-based Image Quality). The algorithm works based on the assumption that the quality of JPEG2000-
compressed images can be evaluated by separately computing the quality of the edge/near-edge regions and
the non-edge regions where no edges are present. EDIQ first separates the input image into edge/near-edge
regions and non-edge regions by applying Canny edge detection and edge-pixel dilation. Our previous sharpness
algorithm, FISH [Vu and Chandler, 2012], is used to generate a sharpness map. The part of the sharpness map
corresponding to the non-edge regions is collapsed by using root mean square to yield the image quality index of
the non-edge regions. The other part of the sharpness map, which corresponds to the edge/near-edge regions, is
weighted by the local RMS contrast and the local slope of magnitude spectrum to yield an enhanced quality map,
which is then collapsed into the quality index of the edge/near-edge regions. These two indices are combined by
a geometric mean to yield a quality indicator of the input image. Testing on the JPEG2000-compressed subsets
of four different image-quality databases demonstrate that EDIQ is competitive with other no-reference image
quality algorithms on JPEG2000-compressed images.