基于改进的 U-Net 卷积神经网络的遥感影像水体信息提取方法-学报杂志社

期刊信息

合肥工业大学（自然科学版），2024年4月，第47卷第4期：488-495，515

DOI: 10.3969/j.issn.1003-5060.2024.04.009

作者信息

宋子俊 $ ^{1,2} $，董张玉 $ ^{1,2,3} $，张鹏飞 $ ^{1,2} $，张远南 $ ^{1,2} $

(1. 合肥工业大学计算机与信息学院, 安徽合肥 230601; 2. 工业安全与应急技术安徽省重点实验室, 安徽合肥 230601; 3. 合肥工业大学智能互联系统安徽省实验室, 安徽合肥 230601)

摘要和关键词

摘要: 针对当前遥感影像水体信息提取存在细节水体提取能力较弱、重要特征损失较大的问题,文章提出一种基于改进的 U-Net 网络实现遥感影像水体信息提取的方法。该方法首先通过引入 Resnet 残差卷积模块深化传统 U-Net 网络架构提升特征挖掘能力,并引入 Respath 残差连接模块减少跳跃连接过程中的语义差距,同时引入 PSConv 多尺度卷积模块、Eca 有效通道注意力机制模块,提高网络特征学习能力,构建 PS-Eca-Multiresunet 网络模型,弥补传统 U-Net 网络存在的细节特征提取能力较弱问题。选择“2020 年第四届中科星图杯高分遥感图像解译软件大赛”数据集进行实验,结果表明,与传统 U-Net 网络模型相比,该方法水体提取的平均交并比提高了 9.08,像素精度提升了 7.4%。改进的网络提取结果能够有效避免阴影影响,提高对细节水体的提取精度,实现遥感影像水体信息的高精度提取。

关键词: 水体提取；深度学习；多尺度卷积；有效通道注意力机制；Multiresunet 网络

Authors

SONG Zijun $ ^{1,2} $, DONG Zhangyu $ ^{1,2,3} $, ZHANG Pengfei $ ^{1,2} $, ZHANG Yuannan $ ^{1,2} $

(1. School of Computer Science and Information Engineering, Hefei University of Technology, Hefei 230601, China; 2. Anhui Province Key Laboratory of Industry Safety and Emergency Technology, Hefei 230601, China; 3. Intelligent Interconnected Systems Laboratory of Anhui Province, Hefei University of Technology, Hefei 230601, China)

Abstract and Keywords

Abstract: Aiming at the problems of weak ability to extract detailed water bodies and large loss of important features in current water body information extraction from remote sensing images, this paper proposes a method to extract water body information from remote sensing images using an improved U-Net network. The method firstly deepens the traditional U-Net network architecture by introducing the Resnet residual convolution module to improve the feature mining ability, and introduces the Res-path residual connection module to reduce the semantic gap in the skip connection process, while introducing the PSConv multi-scale convolution module and Eca effective channel attention mechanism module to improve the network feature learning ability, and constructs the PS-Eca-Multiresunet network model to compensate for the shallow feature loss problem that exists in general networks. The dataset of 2020 GEOVIS Cup Gaofen Challenge on Automated High-Resolution Earth Observation Image Interpretation is selected for the experiment. The results show that the average intersection ratio of water extraction by this method is 9.08 higher than that of the traditional U-Net network model.

Keywords: water body extraction; deep learning; multi-scale convolution; effective channel attention mechanism; Multiresunet network

基金信息

安徽省重点研究与开发计划资助项目（202004a07020030）；安徽省自然科学基金资助项目（2108085MF233）和中央高校基本科研业务费专项资金资助项目（JZ2021HGTB0111）

合肥工业大学学报(自然科学版)

导航菜单

基于改进的 U-Net 卷积神经网络的遥感影像水体信息提取方法