An intelligent wireless channel corrupted image-denoising framework using symmetric convolution-based heuristic assisted residual attention network DOI

S. Pushpa Mala,

Aparna Kukunuri

Network Computation in Neural Systems, Год журнала: 2024, Номер unknown, С. 1 - 34

Опубликована: Май 14, 2024

Image denoising is one of the significant approaches for extracting valuable information in required images without any errors. During process image transmission wireless medium, a wide variety noise presented to affect quality. For efficient analysis, an effective approach needed enhance quality images. The main scope this research paper correct errors and remove effects channel degradation. A corrupted developed channels eliminate bugs. are gathered from at receiver end. Initially, collected decomposed into several regions using Adaptive Lifting Wavelet Transform (ALWT) then "Symmetric Convolution-based Residual Attention Network (SC-RAN)" employed, where residual obtained by separating clean noisy parameters present optimized Hybrid Energy Golden Tortoise Beetle Optimizer (HEGTBO) maximize efficiency. performed over get final denoised numerical findings model attain 31.69% regarding PSNR metrics. Thus, analysis shows improvement.

Язык: Английский

Advances in Deep Learning for Medical Image Analysis: A Comprehensive Investigation DOI
Rajeev Ranjan Kumar, S. Vishnu Shankar, Ronit Jaiswal

и другие.

Journal of Statistical Theory and Practice, Год журнала: 2025, Номер 19(1)

Опубликована: Янв. 23, 2025

Язык: Английский

Процитировано

2
Overview of Research on Digital Image Denoising Methods DOI Creative Commons
Jing Mao,

Lianming Sun,

Jie Chen

и другие.

Sensors, Год журнала: 2025, Номер 25(8), С. 2615 - 2615

Опубликована: Апрель 20, 2025

During image collection, images are often polluted by noise because of imaging conditions and equipment limitations. Images also disturbed external during compression transmission, which adversely affects consequent processing, like segmentation, target recognition, text detection. A two-dimensional amplitude is one the most common categories, widely used in people’s daily life work. Research on this kind image-denoising algorithm a hotspot field denoising. Conventional denoising methods mainly use nonlocal self-similarity sparser representatives converted domain for In particular, three-dimensional block matching filtering (BM3D) not only effectively removes but better retains detailed information image. As artificial intelligence develops, deep learning-based method has become an important research direction. This review provides general overview comparison traditional neural network-based methods. First, essential framework classic network approaches presented, classified summarized. Then, existing compared with quantitative qualitative analyses public dataset. Finally, we point out some potential challenges directions future can help researchers clearly understand differences between various algorithms, helps them to choose suitable algorithms or improve innovate basis ideas subsequent field.

Язык: Английский

Процитировано

0
Ernet: A Deep Framework for Detection and Classification of Lung Cancer from Histopathological Images DOI

Prem Chand Yadava,

Subodh Srivastava

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0
Design of optimized fourth order PDE filter for restoration and enhancement of Microbiopsy images of breast Cancer DOI
Sonam Tyagi, Subodh Srivastava, Bikash Chandra Sahana

и другие.

Multimedia Tools and Applications, Год журнала: 2024, Номер unknown

Опубликована: Июнь 8, 2024

Язык: Английский

Процитировано

1
A Modified of Fourth-Order Partial Differential Equations Model Based on Isophote Direction to Noise Image Removal DOI Creative Commons

Zahra R. Jawad,

Ahmed K. Al-Jaberi

Journal of education for pure science., Год журнала: 2024, Номер 14(3)

Опубликована: Сен. 1, 2024

Image denoising is one of the initial stages image processing. Many models based on diffusion method have been used to smooth image. One problems, we face in model its possible loss edges. The force known be more effective areas high frequency. So This paper suggests combining direction isophote and fourth-order partial differential equations reduce problem edges preserve important details can regulate diffusion. Thus, a proposed that remove noise area while preserving We proven efficiency superiority by applying it set images solving numerically using finite difference (FDM).

Язык: Английский

Процитировано

1
Hybrid Despeckling for Ultrasound Images Using Sticks Filter and Fourth-Order PDE for Enhanced Diagnostic Precision DOI

Jai Jaganath Babu Jayachandran,

M. Rohith,

Lavanya Krishnan

и другие.

Deleted Journal, Год журнала: 2024, Номер 3(5), С. 1 - 8

Опубликована: Ноя. 30, 2024

Speckle noise in ultrasound imaging poses significant challenges by degrading image quality and affecting diagnostic precision. This study evaluates compares the performance of established despeckling algorithms, including Lee, Kuan, Frost, Non-Local Means, PMAD filters, as well advanced techniques such Fourth-Order Partial Differential Equations (PDEs) a novel hybrid method combining Sticks filters with PDE. Quantitative assessment was performed using metrics Peak Signal-to-Noise Ratio (PSNR), Mean Squared Error (MSE), Equivalent Number Looks (ENL), Structural Similarity Index (SSI), Signal-to-Mean Power (SMPI), computational efficiency. Among evaluated methods, Lee filter achieved highest PSNR 25.05 dB, demonstrating effective suppression while preserving details image. The combination PDE ENL 0.0331, indicating superior smoothing homogeneous regions enhanced contrast. While exhibited speckle minimal MSE 886.49, it introduced slight blurring, compromising structural details. Visual inspections revealed that approach delivered exceptional edge preservation contrast enhancement, outperforming other clinical scenarios thyroid nodule analysis. results demonstrate proposed addresses critical trade-offs between detail preservation, offering robust framework to improve utility images. Future research could explore optimizing these algorithms for real-time applications, enabling broader adoption.

Язык: Английский

Процитировано

1
An intelligent wireless channel corrupted image-denoising framework using symmetric convolution-based heuristic assisted residual attention network DOI

S. Pushpa Mala,

Aparna Kukunuri

Network Computation in Neural Systems, Год журнала: 2024, Номер unknown, С. 1 - 34

Опубликована: Май 14, 2024

Image denoising is one of the significant approaches for extracting valuable information in required images without any errors. During process image transmission wireless medium, a wide variety noise presented to affect quality. For efficient analysis, an effective approach needed enhance quality images. The main scope this research paper correct errors and remove effects channel degradation. A corrupted developed channels eliminate bugs. are gathered from at receiver end. Initially, collected decomposed into several regions using Adaptive Lifting Wavelet Transform (ALWT) then "Symmetric Convolution-based Residual Attention Network (SC-RAN)" employed, where residual obtained by separating clean noisy parameters present optimized Hybrid Energy Golden Tortoise Beetle Optimizer (HEGTBO) maximize efficiency. performed over get final denoised numerical findings model attain 31.69% regarding PSNR metrics. Thus, analysis shows improvement.

Язык: Английский

Процитировано

0