MDLatLRR: A Novel Decomposition Method for Infrared and Visible Image Fusion

By Hui Li, Xiaojun Wu, Josef Kittler. IEEE Transactions on Image Processing, 2018

543 citations recorded by the source corpus.

Image decomposition is crucial for many image processing tasks, as it allows to extract salient features from source images. A good image decomposition method could lead to a better performance, especially in image fusion tasks. We propose a multi-level image decomposition method based on latent low-rank representation(LatLRR), which is called MDLatLRR. This decomposition method is applicable to many image processing fields. In this paper, we focus on the image fusion task. We build a novel image fusion framework based on MDLatLRR which is used to decompose source images into detail parts(salient features) and base parts. A nuclear-norm based fusion strategy is used to fuse the detail parts and the base parts are fused by an averaging strategy. Compared with other state-of-the-art fusion methods, the proposed algorithm exhibits better fusion performance in both subjective and objective evaluation.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/c84c4816ae47fb319151a86c7a8c571233c04799

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Hui Li
2. Yiannis Demiris
3. Xingchen Zhang
4. Jiayi Ma
5. Xiaojun Wu
6. Linfeng Tang
7. Josef Kittler
8. Y. Liu

In Media Technology

1. Hui Li
2. Yiannis Demiris
3. Xingchen Zhang
4. Linfeng Tang
5. Jiayi Ma
6. Xiaojun Wu
7. Josef Kittler
8. Wei Tang

Citations in the corpus

• LRRNet: A Novel Representation Learning Guided Fusion Network for Infrared and Visible Images (2023, methodology): “As discussed in the review of LRR-based fusion methods [11], LRR is a powerful representation model for image decomposition.”
• SuperFusion: A Versatile Image Registration and Fusion Network with Semantic Awareness (2022, unknown): “, multi-scale transformation (MST)based [8], [9], [10], sparsity representation (SR)-based [11], subspace-based [12], saliency-based [13], and optimizationbased [14], [15] fusion methods.”
• RFN-Nest: An end-to-end residual fusion network for infrared and visible images (2021, methodology): “In the feature extraction direction, deep learning meth-ods are utilized to extract deep representation of the information conveyed by the source images [22][23][24][25][26].”
• DATFuse: Infrared and Visible Image Fusion via Dual Attention Transformer (2023, methodology): “29], AUIF [1], DenseFuse [53], FusionGAN [6], GANMcC [55], RFN-Nest [54], CSF [56], MFEIF [24], PPTFusion [67], SwinFuse [7], and the proposed method.”
• STDFusionNet: An Infrared and Visible Image Fusion Network Based on Salient Target Detection (2021, methodology): “…and visible image fusion algorithms have been proposed, which could be broadly classified into five classes, including multiscale decomposition-based methods [2], [3], saliency-based methods [4], sparse representation-based methods [5], optimization-based methods [6], and hybrid-based meth-ods [7].”
• UIU-Net: U-Net in U-Net for Infrared Small Object Detection (2022, unknown): “deep learning methods for multi-source data fusion [53], [54] will be investigated in future work to improve the performance of object detection.”
• Visible and Infrared Image Fusion Using Deep Learning (2023, background): “active research topic for many years [1], [2], [3], [4], [5], [6].”
• NestFuse: An Infrared and Visible Image Fusion Architecture Based on Nest Connection and Spatial/Channel Attention Models (2020, methodology): “These fused features are fed into the nest connection-based decoder network to generate the fused image.”