TY - GEN
T1 - Machine Learning for Image Denoising
T2 - International Conference on Medical Imaging and Computer-Aided Diagnosis, MICAD 2023
AU - Wu, Meng
AU - Wang, Shuihua
AU - Chen, Shuwen
AU - Zhang, Yudong
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.
PY - 2024
Y1 - 2024
N2 - Due to the increasing demand for high-quality images in various applications, more attention is paid to image denoising to improve image quality. However, the traditional image-denoising methods have great limitations for complex noise patterns. Machine learning (ML), especially deep learning (DL), has been given attention to solving this problem and has a good prospect for image denoising. This paper raises the necessity of image denoising first, then some common noise types are introduced, and finally, the technology of denoising— ML is discribed. The combined use of image-denoising methods can figure out complex situations to a large extent, providing a guarantee for high-quality images. Since DL can automatically learn complex patterns and levels from data, DL architectures such as CNNs and GANs are used to denoise. We can determine the quality of denoising results by evaluation metrics that provide quantitative measures, such as PSNR and MSE. This paper comprehensively considers the effectiveness of ML and DL in image denoising, and affirms its future potential.
AB - Due to the increasing demand for high-quality images in various applications, more attention is paid to image denoising to improve image quality. However, the traditional image-denoising methods have great limitations for complex noise patterns. Machine learning (ML), especially deep learning (DL), has been given attention to solving this problem and has a good prospect for image denoising. This paper raises the necessity of image denoising first, then some common noise types are introduced, and finally, the technology of denoising— ML is discribed. The combined use of image-denoising methods can figure out complex situations to a large extent, providing a guarantee for high-quality images. Since DL can automatically learn complex patterns and levels from data, DL architectures such as CNNs and GANs are used to denoise. We can determine the quality of denoising results by evaluation metrics that provide quantitative measures, such as PSNR and MSE. This paper comprehensively considers the effectiveness of ML and DL in image denoising, and affirms its future potential.
KW - Deep learning architectures
KW - Evaluation metrics
KW - Machine learning
UR - http://www.scopus.com/inward/record.url?scp=85188675390&partnerID=8YFLogxK
U2 - 10.1007/978-981-97-1335-6_30
DO - 10.1007/978-981-97-1335-6_30
M3 - Conference Proceeding
AN - SCOPUS:85188675390
SN - 9789819713349
T3 - Lecture Notes in Electrical Engineering
SP - 340
EP - 351
BT - Proceedings of 2023 International Conference on Medical Imaging and Computer-Aided Diagnosis (MICAD 2023) - Medical Imaging and Computer-Aided Diagnosis
A2 - Su, Ruidan
A2 - Zhang, Yu-Dong
A2 - Frangi, Alejandro F.
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 9 December 2023 through 10 December 2023
ER -