CT and MRI image fusion based on multiscale decomposition method and hybrid approach

IET Image Processing, Journal Year: 2018, Volume and Issue: 13(1), P. 83 - 88

Published: Oct. 12, 2018

In the fusion process of medical computed tomography (CT) and magnetic resonance image (MRI), traditional multiscale methods often reduce contrast fused images. Although sparse representation (SR) overcome this shortcoming, they are too smooth along strong edges image. To these shortcomings, CT MRI based on decomposition method hybrid approach is proposed. There three main steps. First, cartoon parts texture obtained by improved using global gradients. Second, large structure specific dictionary ‘L1‐max norm’ principle. The textured non‐subsampled contourlet transformation (NSCT) maximum energy rule. Finally, final result superimposing part part. experimental results demonstrate that proposed outperforms state‐of‐the‐art SR NSCT in terms visual effect objective quality.

Language: Английский

---

Segmentation of scanning tunneling microscopy images using variational methods and empirical wavelets

Pattern Analysis and Applications, Journal Year: 2019, Volume and Issue: 23(2), P. 625 - 651

Published: May 5, 2019

Language: Английский

---

A Non-Local Dual-Domain Approach to Cartoon and Texture Decomposition

IEEE Transactions on Image Processing, Journal Year: 2018, Volume and Issue: 28(4), P. 1882 - 1894

Published: Nov. 19, 2018

This paper addresses the problem of cartoon and texture decomposition. Microtextures are characterized by their power spectrum, we propose to extract components from information provided spectrum image patches. The contribution a patch is detected as statistically significant spectral with respect null hypothesis modeling non-textured patch. null-hypothesis model built upon coarse representation obtained basic yet fast filtering algorithm literature. Hence, term "dual domain": decomposition in spatial domain an input proposed approach. statistical also patches similar textures across image. approach, therefore, falls within family non-local methods. Experimental results shown various application areas, including canvas pattern removal fine arts painting, or periodic noise remote sensing imaging.

Language: Английский

---

Image-based haptic display via a novel pen-shaped haptic device on touch screens

Multimedia Tools and Applications, Journal Year: 2017, Volume and Issue: 76(13), P. 14969 - 14992

Published: Jan. 21, 2017

Language: Английский

---

Content-Based Image Retrieval using Local Binary Curvelet Co-occurrence Pattern—A Multiresolution Technique

The Computer Journal, Journal Year: 2017, Volume and Issue: 61(3), P. 369 - 385

Published: Sept. 2, 2017

With the growth of various image-capturing devices, image acquisition is no longer a difficult task. As this technology flourishing, types complex images are being produced. In order to access large number stored in database easily, must be properly organized. Field retrieval attempts solve problem. produced, processing them using single-resolution techniques not sufficient as these may contain varying levels details. This paper proposes novel multiresolution descriptor, local binary curvelet co-occurrence pattern, achieve task content-based retrieval. Curvelet transform grayscale computed followed by computation pattern resulting coefficients. Finally, feature vector constructed grey-level matrix which matched with images. The proposed descriptor combines properties and technique transform, efficiently covers curvilinear geometrical structures present image. Performance method measured terms precision recall tested on five benchmark datasets consisting natural has been compared single well some other state-of-the-art methods. experimental results clearly demonstrate that produces high accuracy outperforms recall.

Language: Английский

---

Blind Quality Metric for Multidistortion Images Based on Cartoon and Texture Decomposition

IEEE Signal Processing Letters, Journal Year: 2016, Volume and Issue: 23(9), P. 1265 - 1269

Published: July 27, 2016

In this letter, a no-reference (NR) hybrid image quality assessment (IQA) metric based on cartoon-texture decomposition (CTD) is presented. Focusing images distorted by both blur and noise, the method takes properties of CTD to separate into cartoon part with salient edges texture noises. Then, degree noise level can be estimated separately from different parts, combine joint effect prediction between distortions, we present decomposition-based blind (CTDBBM). Comparative studies classical full-reference IQA metrics state-of-the-art NR are conducted multidistortion database: LIVEMD. Experimental results show that CTDBBM performs well has high consistency human opinions given in database.

Language: Английский

---

Image decomposition fusion method based on sparse representation and neural network

Applied Optics, Journal Year: 2017, Volume and Issue: 56(28), P. 7969 - 7969

Published: Sept. 28, 2017

For noisy images, in most existing sparse representation-based models, fusion and denoising proceed simultaneously using the coefficients of a universal dictionary. This paper proposes an image method based on cartoon + texture dictionary pair combined with deep neural network combination (DNNC). In our model, are carried out alternately. The proposed is divided into three main steps: denoising. More specifically, (1) denoise source images external/internal methods separately; (2) fuse these preliminary denoised results to obtain external representation result (E-CTSR) internal (I-CTSR); (3) combine E-CTSR I-CTSR DNNC (EI-CTSR) final result. Experimental demonstrate that EI-CTSR outperforms not only stand-alone but also state-of-the-art such as (SR) adaptive (ASR) for isomorphic SR ASR heterogeneous multi-mode images.

Language: Английский

---

An Image Decomposition Fusion Method for Medical Images

Mathematical Problems in Engineering, Journal Year: 2020, Volume and Issue: 2020, P. 1 - 11

Published: July 29, 2020

A fusion method based on the cartoon+texture decomposition and convolution sparse representation theory is proposed for medical images. It can be divided into three steps: firstly, cartoon texture parts are obtained using improved cartoon-texture method. Secondly, rules of energy protection feature extraction used in part, while part. Finally, fused image superimposing parts. Experiments show that algorithm effective.

Language: Английский

---

No-Reference Quality Assessment Method for Blurriness of SEM Micrographs with Multiple Texture

Scanning, Journal Year: 2019, Volume and Issue: 2019, P. 1 - 15

Published: June 2, 2019

Scanning electron microscopy (SEM) plays an important role in the intuitive understanding of microstructures because it can provide ultrahigh magnification. Tens or hundreds images are regularly generated and saved during a typical imaging process. Given subjectivity microscopist's focusing operation, blurriness is distortion that debases quality micrographs. The selection high-quality micrographs using subjective methods expensive time-consuming. This study proposes new no-reference assessment method for evaluating SEM human visual system more sensitive to distortions cartoon components than those redundant textured according Gestalt perception psychology entropy masking property. Micrographs initially decomposed into components. Then, spectral spatial sharpness maps extracted. One metric calculated by combining other on basis edge maximum local variation map Finally, two metrics combined as final metric. objective scores this exhibit high correlation consistency with scores.

Language: Английский

---

W-LDMM: A Wasserstein driven low-dimensional manifold model for noisy image restoration

Neurocomputing, Journal Year: 2019, Volume and Issue: 371, P. 108 - 123

Published: Sept. 6, 2019

Language: Английский

---