Comparison of neural networks for suppression of multiplicative noise in images

被引：0

作者：

Pavlov, V. A. ^{[1
]}

Belov, A. A. ^{[1
]}

Nguen, V. T. ^{[1
]}

Jovanovski, N. ^{[1
]}

Ovsyannikova, A. S. ^{[1
]}

机构：

[1] Peter Great St Petersburg Polytech Univ, Inst Elect & Telecommun, Polytech Skaya 29, St Petersburg 195251, Russia

来源：

COMPUTER OPTICS | 2024年 / 48卷 / 03期

关键词：

speckle noise; radar image; SAR; noise reduction; image processing; neural network;

D O I：

10.18287/2412-6179-CO-1400

中图分类号：

O43 [光学];

学科分类号：

070207 ; 0803 ;

摘要：

The paper compares several neural network (NN) architectures for suppression of multiplicative noise. The images may contain sharp boundaries and large homogeneous areas. Convolutional and fully connected networks are investigated. It is shown that different architectures require significantly different amount of training data to reach the same noise suppression quality. Examples of NN requiring lower amounts of training data are presented.

引用

页码：425 / 432

页数：8