首页|基于深度学习的高光谱图像去噪综述

基于深度学习的高光谱图像去噪综述

扫码查看
高光谱图像具有图谱合一的优点,已被广泛应用于农业、地球科学和地质灾害等领域。由于噪声的影响往往限制了高光谱图像的应用,高光谱图像去噪已成为一种重要的图像预处理方式。深度学习作为近些年来快速发展的技术之一,已被成功地应用于高光谱图像去噪中。基于深度学习的高光谱图像去噪研究成果正逐年增加,为了便于对该领域进行更系统全面的研究,本文概述了基于深度学习的高光谱图像去噪研究进展,对现有主要研究成果进行了分类、归纳与总结,并对该领域的未来研究趋势进行了展望。
A review of hyperspectral image denoising based on deep learning
Hyperspectral images are widely used in agriculture,earth science,geological disasters and other fields because hyperspectral images have the advantage that image integrates with spectrum.However,in the imaging process of hyperspec-tral image,due to various factors,hyperspectral images will inevitably be polluted by multiple noises.The existence of noises often limits the application value of hyperspectral images.Therefore,hyperspectral image denoising is an important way of image preprocessing.As one of the rapidly developing technologies in recent years,deep learning has been successfully ap-plied in the task of hyperspectral image denoising.The number of research results on deep learning-based hyperspectral im-age denoising is increasing year by year.In order to facilitate a more systematic and comprehensive exploration of this field,this paper outlines the research progress of deep learning-based hyperspectral image denoising,classifies and summarizes the existing main research findings.Finally,the future research trends in this field are prospected.

hyperspectral image denoisingdeep learningconvolution neural networkrecurrent neural networkattention mechanism

张俊、谭耀鑫、卢静静、徐晨光、邓承志

展开 >

南昌工程学院江西省水信息协同感知与智能处理重点实验室,江西南昌 330099

南昌工程学院工程数学与先进计算重点实验室,江西南昌 330099

高光谱图像去噪 深度学习 卷积神经网络 循环神经网络 注意力机制

江西省自然科学基金资助项目南昌工程学院研究生创新计划项目

20232BAB201017YJSCX202316

2024

南昌工程学院学报
南昌工程学院

南昌工程学院学报

影响因子:0.272
ISSN:1006-4869
年,卷(期):2024.43(3)
  • 5