基于双通道特征增强的高光谱图像分类  

Hyperspectral Image Classification Based on Dual-Channel Feature Enhancement

作　　者：赵利[1] 王雷全[1] 张俊三[1] 邵志敏 朱杰[3] Zhao Li;Wang Leiquan;Zhang Junsan;Shao Zhimin;Zhu Jie(College of Computer Science and Technology,China University of Petroleum(East China),Qingdao 266580,Shandong,China;State Grid Shandong Electric Power Company,Jinan 250003,Shandong,China;Department of Information Management,the National Police University for Criminal Justice,Baoding 071000,Hebei,China)

机构地区：[1]中国石油大学(华东)计算机科学与技术学院,山东青岛266580 [2]国网山东省电力公司,山东济南250003 [3]中央司法警官学院信息管理系,河北保定071000

出　　处：《激光与光电子学进展》2023年第12期138-149,共12页Laser & Optoelectronics Progress

基　　金：山东省自然科学基金(ZR2020MF006);中国石油大学(华东)自主创新科研计划项目(20CX05019A)。

摘　　要：针对如何在训练样本有限的情况下更加充分提取和利用高光谱图像的空间信息和光谱信息这一问题,提出一种基于双通道特征增强(DCFE)的高光谱图像分类方法。首先,设计两个通道分别捕获光谱特征和空间特征,在每个通道中使用三维卷积作为特征提取器。然后,将降维后的光谱通道中的特征图与空间通道的特征图进行融合。最后,将融合了光谱特征和空间特征的特征图输入注意力模块中,通过提升重要信息的关注度和降低无用信息的干扰来实现特征增强。实验结果表明,所提方法在Indian Pines(3%训练样本)、Pavia University(0.5%训练样本)、Salinas(0.5%训练样本)和Botswana(1.2%训练样本)等4个高光谱数据集上的总体分类精度分别为96.57%、98.15%、98.95%和96.83%,与其他5种高光谱分类方法相比,所提方法在分类性能上取得了明显提升。A classification method of hyperspectral images based on dual channel feature enhancement(DCFE)is proposed to solve the problem of how to extract and use the spatial and spectral information of hyperspectral images more fully when the training samples are limited.First,two channels are designed to capture spectral and spatial features,and 3D convolution is used as a feature extractor in each channel.The feature map from the reduced-dimension spectral channel is fused with the feature map of the spatial channel.Finally,the feature map combining spectral and spatial features is input into the attention module,and feature enhancement is achieved by increasing attention to important information while decreasing interference from irrelevant information.The experimental results show that the proposed method has an overall classification accuracy of 96.57%,98.15%,98.95%,and 96.83%on four hyperspectral data sets,including Indian Pines(3%training sample),Pavia University(0.5%training sample),Salinas(0.5%training sample),and Botswana(1.2%training sample),respectively.When compared to the other five hyperspectral classification methods,the proposed method has remarkably improved the classification performance.

关 键 词：图像处理 高光谱图像分类 双通道 特征增强 注意力机制 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]