High quality multi-spectral and panchromatic image fusion technologies based on Curvelet transform

Dong, Limin; Yang, Qingxiang; Wu, Houming; Xiao, Huaibao; Xu, Mingliang

doi:10.1016/j.neucom.2015.01.050

Cited by 100 publications

(40 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared with other multiscale decomposition tools, shearlets not only enjoy similar nonlinear order of error as curvelets do when approximating the optimum but also subdivide layer by layer in frequency domain [19] [20]. In addition, both the number of directions for the decomposition and the size of the support basis are unlimited, hence, shearlets have higher efficiency in computation compared with contourlets [21] [22].…”

Section: State Of the Artmentioning

confidence: 99%

Medical Image Fusion Combined with Accelerated Non-negative Matrix Factorization and Expanded Laplacian Energy in Shearlet Domain

Lai¹,

Wang²,

He³

et al. 2017

JESTR

View full text Add to dashboard Cite

Shearlet transform has been widely applied in related fields due to its admirable properties in image approximation. Image fusion method based on accelerated non-negative matrix factorization (ANMF) and expanded energy of Laplace (EEOL) rules was proposed in this study to integrate the complementary information of medical images with multiple modalities and improve the accuracy of clinical diagnosis and therapy. First, the registered medical images were decomposed into low-and high-frequency sub-band coefficients in shearlet domain. Then, the ANMF rule was used in merging low-frequency coefficients. Next, the visual-contrast-based EEOL rule was adopted in extracting details of source images from high-frequency coefficients. Finally, the ultimate fused image was reconstructed by applying inverse shearlet transform. Experimental results reveal that aside from visual effect, the proposed method achieves the best in three of five criteria and the run time is reduced by 29.21% compared with a method based on non-subsampled contourlet transform (NSCT) in computed tomography (CT)-magnetic resonance imaging (MRI) fusion. Moreover, the proposed method takes the first place in four of five criteria with run time reduced by 48.32% and 24.55% compared with two shearlet-based methods in a MRI-positron emission tomography (PET) case. This study indicates that the proposed method is superior to the selected approaches in visual and statistical evaluation, which is conducive to clinical practice of medical image fusion.

show abstract

Section: State Of the Artmentioning

confidence: 99%

Medical Image Fusion Combined with Accelerated Non-negative Matrix Factorization and Expanded Laplacian Energy in Shearlet Domain

Lai¹,

Wang²,

He³

et al. 2017

JESTR

View full text Add to dashboard Cite

show abstract

“…However, IR image typically has lower spatial resolution and fewer details. Image fusion is widely used in remote sensing [4], military [5,6], objects tracking and detecting [7,8], etc. IR and visible image fusion will also bring great benefit to railway clearance intrusion detection.…”

Section: Introductionmentioning

confidence: 99%

Novel Registration and Fusion Algorithm for Multimodal Railway Images with Different Field of Views

Guo

Zhou

Lin

et al. 2018

Journal of Advanced Transportation

View full text Add to dashboard Cite

Objects intruding high-speed railway clearance do great threat to running trains. In order to improve accuracy of railway intrusion detection, an automatic multimodal registration and fusion algorithm for infrared and visible images with different field of views is presented. The ratio of the nearest to next nearest distance, geometric, similar triangle, and RANSAC constraints are used to refine the matching SURF feature points successively. Correct matching points are accumulated with multiframe to overcome the insufficient matching points in single image pair. After being registered, an improved Contourlet transform fusion algorithm combined with total variation and local region energy is proposed. Inverse Contourlet transform to low frequency subband coefficient fused with total variation model and high frequency subband coefficients fused with local region energy is used to reconstruct the fused image. The comparison to other 4 popular fusion methods shows that our algorithm has the best comprehensive performance for multimodal railway image fusion.

show abstract

“…Representing images with a few number of basis or atoms from an over-complete dictionary [15,16] is the core idea of sparse representation theory in image processing. Recent research shows that representing fused information of multiple images from same or different sensor modalities can be achieved with an appropriate sparsity model and well-constructed dictionaries [17,18,19,20,21,22,23,24]. In image fusion, by the difference of dictionaries, sparse representation can be divided into two categories, the fixed-dictionary based sparse representation and dictionary-learning based sparse representation.…”

Section: Introductionmentioning

confidence: 99%

“…In image fusion field, Yang and Li [13] took the first step of applying sparse representation for image fusion. They used a fixed dictionary with DCT basis for sparse representation.Dong and Yang [19] proposed a curvelet transform dictionary based sparse representation for image fusion. They achieved great results in both gray level and colour image fusion.…”

Section: Introductionmentioning

confidence: 99%

A novel dictionary learning approach for multi-modality medical image fusion

et al. 2016

View full text Add to dashboard Cite

High quality multi-spectral and panchromatic image fusion technologies based on Curvelet transform

Cited by 100 publications

References 29 publications

Medical Image Fusion Combined with Accelerated Non-negative Matrix Factorization and Expanded Laplacian Energy in Shearlet Domain

Medical Image Fusion Combined with Accelerated Non-negative Matrix Factorization and Expanded Laplacian Energy in Shearlet Domain

Novel Registration and Fusion Algorithm for Multimodal Railway Images with Different Field of Views

A novel dictionary learning approach for multi-modality medical image fusion

Contact Info

Product

Resources

About