Multi-level contrast filtering in image difference metrics

Simone, Gabriele; Pedersen, Marius; Farup, Ivar; Oleari, Claudio

doi:10.1186/1687-5281-2013-39

Cited by 4 publications

(2 citation statements)

References 31 publications

(42 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each duplicate is separately and simultaneously convolved with two difference of Gaussian filters (DOG). The filters, and are defined as: where and are the widths of the Gaussian filter The motivation behind the use of DOG filter is its efficient application in edge detection for feature enhancement, blob detection, face detection and quality evaluation [ 33 – 36 ]. The DOG filter was implemented using the matlab code available in [ 37 , 38 ].…”

Section: Methodsmentioning

confidence: 99%

Blind blur assessment of MRI images using parallel multiscale difference of Gaussian filters

et al. 2018

Self Cite

View full text Add to dashboard Cite

BackgroundRician noise, bias fields and blur are the common distortions that degrade MRI images during acquisition. Blur is unique in comparison to Rician noise and bias fields because it can be introduced into an image beyond the acquisition stage such as postacquisition processing and the manifestation of pathological conditions. Most current blur assessment algorithms are designed and validated on consumer electronics such as television, video and mobile appliances. The few algorithms dedicated to medical images either requires a reference image or incorporate manual approach. For these reasons it is difficult to compare quality measures from different images and images with different contents. Furthermore, they will not be suitable in environments where large volumes of images are processed. In this report we propose a new blind blur assessment method for different types of MRI images and for different applications including automated environments.MethodsTwo copies of the test image are generated. Edge map is extracted by separately convolving each copy of the test image with two parallel difference of Gaussian filters. At the start of the multiscale representation, the initial output of the filters are equal. In subsequent scales of the multiscale representation, each filter is tuned to different operating parameters over the same fixed range of Gaussian scales. The filters are termed low and high energy filters based on their characteristics to successively attenuate and highlight edges over the range of multiscale representation. Quality score is predicted from the distance between the normalized mean of the edge maps at the final output of the filters.ResultsThe proposed method was evaluated on cardiac and brain MRI images. Performance evaluation shows that the quality index has very good correlation with human perception and will be suitable for application in routine clinical practice and clinical research.

show abstract

Section: Methodsmentioning

confidence: 99%

Blind blur assessment of MRI images using parallel multiscale difference of Gaussian filters

et al. 2018

Self Cite

View full text Add to dashboard Cite

show abstract

“…Johnson and Fairchild [JF01] further improve the filters used in S–CIELAB metric while Wang and Hardeberg adapt S–CIELAB to employ six channels of information (lightness, chroma, hue, compression, noise and sharpness) subject to bilateral filtering [WH09, WH12]. Similarly, Simone et al [SOF09a] incorporate the use of log‐compressed OSA‐UCS space [SOF09b] to adapt S–CIELAB into a new metric called S–DEE. Also, Pedersen and Hardeberg [PH09] incorporate spatial properties of the human visual system in extending a hue‐angle algorithm [HL02] to devise a new S–CIELAB like metric called SHAME.…”

Section: Proposed Methodsmentioning

confidence: 99%

Robust Distribution‐Aware Color Correction for Single‐Shot Images

Dhillon,

Joshi,

Baron

et al. 2023

Computer Graphics Forum

View full text Add to dashboard Cite

Color correction for photographed images is an ill‐posed problem. It is also a crucial initial step towards material acquisition for inverse rendering methods or pipelines. Several state‐of‐the‐art methods rely on reducing color differences for imaged reference color chart blocks of known color values to devise or optimize their solution. In this paper, we first establish through simulations the limitation of this minimality criteria which in principle results in overfitting. Next, we study and propose a few spatial distribution measures to augment the evaluation criteria. Thereafter, we propose a novel patch‐based, white‐point centric approach that processes luminance and chrominance information separately to improve on the color matching task. We compare our method qualitatively with several state‐of‐the art methods using our augmented evaluation criteria along with quantitative examinations. Finally, we perform rigorous experiments and demonstrate results to clearly establish the benefits of our proposed method.

show abstract