1,720,978 research outputs found
SHORT REFERENCE IMAGE QUALITY RATING BASED ON ANGULAR EDGE COHERENCE
No full textA very concise comparison technique for objective image quality assessment is described. The method is based on a angular edge coeherence measure defined by local image expansion into a set of harmonic angular functions. Using the angular edge coherence it is possible to estimate the relative quality of a reproduced image with respect to the original one comparing only the values of a single statistical index disjointly extracted from each image. The present contribution briefly illustrates the mathematical support of this technique and provides some significant experimental examples
A structural coherence approach to full reference image quality assessment
No full textIn this contribution a novel technique for objective full reference quality rating of degraded images using the paradigm of structural distortion measurement is presented. The method is based on the separate evaluation of the edge degradation and of the textural disturbances to measure the quality loss of a given image. Compared with other techniques, the method presents quite uniform behavior with respect to different image defects, responsiveness to image improvements, and diagnostic capabilities
Fast calibrating full reference universal quality meter.
No full textA novel approach to Full Reference objective image quality
measurement is presented. It is based on modelling the human
judgment of image quality as a consequence of the value
assumed by a two-dimensional cognitive state. The method
presented here is immune to the variability of causes of quality
impairment, i.e., it is inherently universal. It is shown in this
contribution that this quality meter is fast calibrating, i.e., it
may be calibrated on the basis of few subjective experimental
data
Structure oriented image quality assessment based on multiple statistics
No full textMultiple statistics are employed for evaluation of the fidelity
reproduction of different structures within a Full Reference
image assessment scheme. Such an approach permits the
diagnostic analysis of the basic factors impairing the quality.
Higher level combination of these statistics yields an overall
objective quality index well correlated with subjective
measurement
Application of Local Fisher Information Analysis to Salient Points Extraction
No full textThe Fisher information of local patterns about localization parameters can be used for selecting interesting points in computer vision applications. In this contribution it is shown that evaluation of local Fisher information about position, rotation and size for every point can be performed through a proper filter bank. It is recognized that this approach leads to an extension of the popular Harris salient point detection method, which actually searches for the patterns characterized by relative maxima of the Fisher information about translation. Preliminary experiments indicate that the Fisher information about rotational and scale parameters can be used to improve the repeatability rates of the salient point selection process with respect to the Harris and wavelet based salient point extraction methods
Equalized structural similarity index for image quality assessment
No full textA novel version of the popular SSIM image quality assessment technique, specifically designed to yield uniform MOS prediction in presence of different types of image distortions is presented. Response equalization is obtained through the polynomial combination of a basic SSIM metric with an auxiliary metric characterized by a diverse sensitivity. Accuracy and consistency of this composite index, called E-SSIM, are demonstrated with experiments conducted on two independent archives annotated with MOS values. © 2011 University of Zagreb
Local orientation estimation by tomographic Hermite slices
No full textIdentifying and estimating the orientation of monodimensional (linear) patterns found into images is an important task for pattern recognition purposes (e.g., in SAR images) and can greatly improve the efficacy of image restoration, sharpening and de-noising procedures. Existing approaches to linear pattern orientation estimation are based either on pyramid filter banks, steered to a small set of discrete orientations, or on parametric approaches based on the tensor gradient. In the present work, using a local tomography paradigm, the complementary properties of Hermite and Gauss-Laguerre image expansions are exploited for accurately estimating the orientation angles of linear patterns by rooting a polynomial, built with transform coefficients at each analysis point of the image. In particular, as in direction finding with sensor arrays (e.g., ROOT-MUSIC, MODE and ESPRIT), rooting allows a fast and accurate orientation estimation on a continuous set of angles. It is shown that the feasibility of this scheme is based on the simple link existing between Hermite and Gauss-Laguerre coefficients
Salient Points Detection of Human Faces by a Circular Symmetry Index Based on Fisher's Information
No full textTwo-dimensional approach to full-reference image quality assessment based on positional structural iInformation
No full textA method for full-reference visual quality assessment based on the 2-D combination of two diverse metrics is described. The first metric is a measure of structural information loss based on the Fisher information about the position of the structures in the observed images. The second metric acts as a categorical indicator of the type of distortion that images underwent. These two metrics constitute the inner state of a virtual cognitive model, viewed as a system whose output is the automatic visual quality estimate. The use of a 2-D metric fills the intrinsic incompleteness of methods based on a single metric while providing consistent response across different image impairment factors and blind distortion classification capability with a modest computational overhead. The high accuracy and robustness of the method are demonstrated through cross-validation experiments
- …
