Article
Geochemistry & Geophysics
Shao Xiang, Qiaokang Liang, Leyuan Fang
Summary: High-ratio image compression is a difficult task for remote sensing images due to their complex backgrounds and weak correlation between features. This study proposes a novel entropy model (DWTGMM) based on discrete wavelet transform (DWT) and Gaussian mixture model (GMM) to enhance the representation ability of compression models and estimate the probability distributions of latent representations. Experimental results show that the proposed method achieves excellent performance in remote sensing image compression.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Mathematics
H. K. Nigam, H. M. Srivastava
Summary: This paper investigates the use of nonlinear diffusion for denoising audio signals by applying it to wavelet coefficients obtained from different filters. The results show a significant improvement in denoising compared to wavelet shrinkage.
Article
Optics
Muhammad Rafiq Abuturab, Ayman Alfalou
Summary: A new method for multiple color image fusion, compression, and encryption using compressive sensing, chaotic-biometric keys, and optical fractional Fourier transform is proposed in this paper. The proposed cryptosystem has advantages of reduced data storage, uniqueness of biometric keys in CBPMs, very sensitive orders of the FrFT, and a single-channel hybrid optoelectronic setup.
OPTICS AND LASER TECHNOLOGY
(2022)
Article
Computer Science, Artificial Intelligence
Zhen Chao, Xingguang Duan, Shuangfu Jia, Xuejun Guo, Hao Liu, Fucang Jia
Summary: Medical image fusion of images obtained via different modes can expand the inherent information of original images, whereby the fused image has a superior ability to display details than the original sub images, to facilitate diagnosis and treatment selection. The proposed method based on the discrete stationary wavelet transform (DSWT) and radial basis function neural network (RBFNN) effectively combines the information and details of two images, resulting in a significantly better performance compared to current state-of-the-art methods.
APPLIED SOFT COMPUTING
(2022)
Article
Geochemistry & Geophysics
Xinlei Jia, Yali Peng, Jun Li, Bao Ge, Yunhong Xin, Shigang Liu
Summary: A dual-complementary convolution network is proposed for repairing noisy remote-sensing images by generating multiresolution inputs to reduce network parameters. Experimental evaluations show that this network outperforms other methods on remote-sensing public datasets.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2022)
Article
Computer Science, Artificial Intelligence
Shiveta Bhat, Deepika Koundal
Summary: The study aims to improve the quality and efficiency of Multi-focus Image Fusion (MFIF) technique, proposing a hybrid method based on Neutrosophic Set and Stationary Wavelet Transform. Through quantitative and qualitative evaluations on two different datasets, it is found that the NSWT technique performs well both quantitatively and qualitatively, aiding in extending the Depth-of-Field of the imaging system.
APPLIED SOFT COMPUTING
(2021)
Review
Computer Science, Artificial Intelligence
Zhaobin Wang, Yikun Ma, Yaonan Zhang
Summary: This paper provides a comprehensive review of deep learning-based methods for fusing remote sensing images at the pixel level, discussing the research progress and current status. It suggests that generative adversarial networks have great potential in image generation and unsupervised learning, and the fusion of radar and optical images at the pixel level requires more attention in future work.
INFORMATION FUSION
(2023)
Article
Environmental Sciences
Congcong Wang, Wenbin Sun, Deqin Fan, Xiaoding Liu, Zhi Zhang
Summary: A deep learning-based change detection method for high-resolution remote sensing images is proposed, utilizing 2D discrete wavelet transform and adaptive feature weighted fusion to accurately detect changed objects of different scales and reconstruct change maps with clear boundaries. The method effectively reduces false changes and achieves improved detection performance, as validated in quantitative, qualitative, and efficiency analyses compared to state-of-the-art methods.
Article
Engineering, Mechanical
Wei Zhou, Zhongren Feng, Y. F. Xu, Xiongjiang Wang, Hao Lv
Summary: A new signal decomposition method called empirical Fourier decomposition (EFD) is proposed to address the issues of mode mixing and inconsistency in existing methods. Numerical and experimental validations demonstrate that EFD can provide more accurate and consistent decomposition results for signals with multiple non-stationary modes and closely-spaced modes.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Computer Science, Information Systems
N. H. Kaplan, I Erer
Summary: A scale aware enhancement method based on rolling guidance is proposed for remote sensed images in order to improve contrast and edge information while preserving original radiance values. Comparative study shows that the proposed method has significant improvements in contrast gain and enhancement measurement compared to traditional methods.
JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION
(2021)
Article
Mathematical & Computational Biology
Song Yang, Huibin Wang, Hongmin Gao, Lili Zhang
Summary: Convolutional neural networks (CNNs) have achieved excellent performance in object classification and recognition. However, due to limited labeled samples and the specific nature of geographic data, the practical application of CNN methods in remote sensing (RS) image processing is restricted. To address the issue of small sample RS image classification, a discrete wavelet-based multi-level deep feature fusion method is proposed. This method effectively integrates deep feature information from different frequency bands, resulting in low-dimensional features with good discrimination. Experimental results on four benchmark datasets demonstrate the outstanding performance of the proposed method, particularly with limited training samples of one or two per class.
MATHEMATICAL BIOSCIENCES AND ENGINEERING
(2023)
Article
Computer Science, Information Systems
Yuhan Huang, Jiacheng Lu, Nianzhe Chen, Hui Ding, Yuanyuan Shang
Summary: The development of deep learning has improved image inpainting performance. Different tasks require different models. A two-stage inpainting method combining frequency and spatial domain information is proposed. The impacts of commonly used loss functions on structured and textured images are discussed.
MULTIMEDIA SYSTEMS
(2023)
Article
Computer Science, Information Systems
Khaldi Amine, Kafi Med Redouane, Moad Med Sayah
Summary: This work proposes a reliable and blind watermarking method for securing medical images exchanged in telemedicine. Medical image watermarking enables accurate patient identification, prevents scan confusion, and minimizes the risk of diagnostic mistakes. The suggested approaches effectively retain a considerable quality of watermarked images, according to experimental results on imperceptibility and robustness, and are resistant against various common attacks.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Samah Alshathri, Ezz El-Din Hemdan
Summary: This paper presents an efficient audio watermarking scheme using wavelet-based image fusion, Arnold transforms, and singular value decomposition (SVD) for secure transmission of medical images and reports in the Medical Internet of Things (MIoT) applications. The scheme combines two medical watermarks into a fused watermark to increase the payload of the inserted medical images and utilizes Arnold transform for watermark scrambling. The results show that the proposed scheme increases capacity and security of implanted medical images transmission without affecting audio signal quality.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Yue Zhang, Zhenfang Liu, Min Huang, Qibing Zhu, Bao Yang
Summary: A multi-resolution depth image restoration method is proposed to improve the edge details of depth images by applying joint bilateral filtering and color-guided filtering, effectively reducing the additive Gaussian noise and decreasing the time consumption of depth image restoration.
MACHINE VISION AND APPLICATIONS
(2021)
