A similarity metric for edge images

Prieto, Miguel S.; Allen, Alastair Robert

doi:10.1109/tpami.2003.1233900

Cited by 91 publications

(48 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…1.1 and 3.8). The requirement of perfect image registration is lifted for the pixel correspondence metric, 42 closest distance metric, 43 or pointto-closest-point MSE, which account for edge distances. Natural images can be handled by the complex wavelet CW-SSIM index but mostly small translations can be supported (Ref.…”

Section: Discussionmentioning

confidence: 99%

A transformation-aware perceptual image metric

2015

View full text Add to dashboard Cite

Abstract. Predicting human visual perception has several applications such as compression, rendering, editing, and retargeting. Current approaches, however, ignore the fact that the human visual system compensates for geometric transformations, e.g., we see that an image and a rotated copy are identical. Instead, they will report a large, false-positive difference. At the same time, if the transformations become too strong or too spatially incoherent, comparing two images gets increasingly difficult. Between these two extrema, we propose a system to quantify the effect of transformations, not only on the perception of image differences but also on saliency and motion parallax. To this end, we first fit local homographies to a given optical flow field, and then convert this field into a field of elementary transformations, such as translation, rotation, scaling, and perspective. We conduct a perceptual experiment quantifying the increase of difficulty when compensating for elementary transformations. Transformation entropy is proposed as a measure of complexity in a flow field. This representation is then used for applications, such as comparison of nonaligned images, where transformations cause threshold elevation, detection of salient transformations, and a model of perceived motion parallax. Applications of our approach are a perceptual level-of-detail for real-time rendering and viewpoint selection based on perceived motion parallax. © The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

show abstract

Section: Discussionmentioning

confidence: 99%

A transformation-aware perceptual image metric

2015

View full text Add to dashboard Cite

show abstract

“…The paper [19] introduces a representation based on a recursive division of a shape. Some methods operate on edge images [30] or contours of objects [23]. Others take advantage of curvature trees [2], contour point histogram [34], deformable potential of contour [38] or curve adjustment procedure based on their minimal deformation cost [39].…”

Section: Previous Workmentioning

confidence: 99%

Tree-based binary image dissimilarity measure with meta-heuristic optimization

Zieliński

Iwanowski

2015

Pattern Anal Applic

View full text Add to dashboard Cite

In this paper, we present a method of evaluating binary image dissimilarity based on tree representation and heuristic optimization. Starting from the image, a graph structure of a binary tree is constructed that splits the set of image foreground pixels into consecutive subsets attached to tree nodes. Next, instead of comparing two images themselves, one compares the trees and expresses image dissimilarity as tree dissimilarity, which can be characterized by a nonlinear function. The goal is to find its minimum, as it corresponds with the best match of compared trees. Searching for the minimum would be ineffective with analytical optimization methods. Hence, we have approached the issue with three meta-heuristic algorithms, namely genetic algorithm, particle swarm optimization (PSO) and simulated annealing. The presented results show that PSO achieved the best results. The proposed method is compared with other binary image comparison approaches. The performed tests that are described in the paper show that it outperforms its competitors and can be successfully applied to compare binary images.

show abstract

“…The peak signal to noise ratio (PSNR) is the most widely used measure of image fidelity. Although PSNR is not suitable to measure precision [27], it is appropriate to measure robustness. The edge detector is applied to both the noiseless and the noisy version of the same image.…”

Section: Robustnessmentioning

confidence: 99%