融合深度特征的高光谱图像小样本分类方法

doi:10.13474/j.cnki.11-2246.2025.1106

测绘通报 ›› 2025, Vol. 0 ›› Issue (11): 34-39.doi: 10.13474/j.cnki.11-2246.2025.1106

融合深度特征的高光谱图像小样本分类方法

秦进春^1,2, 裴航³, 刘冰³, 余岸竹³, 陈俊铭³, 范俊忆³

1. 智能空间信息国家级重点实验室, 北京 100029;
2. 西安测绘研究所, 陕西西安 710054;
3. 信息工程大学地理空间信息学院, 河南郑州 450001

收稿日期:2025-03-20 发布日期:2025-12-04
通讯作者: 裴航。E-mail:peihang315@163.com
作者简介:秦进春(1989—),男,博士,助理研究员,主要研究方向为多源遥感数据智能处理与应用。E-mail:qjc20@tsinghua.org.cn
基金资助:
国家自然科学基金(42401501)

Few-shot hyperspectral image classification with depth feature fusion

QIN Jinchun^1,2, PEI Hang³, LIU Bing³, YU Anzhu³, CHEN Junming³, FAN Junyi³

1. National Key Laboratory of Intelligent Spatial Information, Beijing 100029, China;
2. Xi'an Research Institute of Surveying and Mapping, Xi'an 710054, China;
3. Institute for Geospatial Information, Information Engineering University, Zhengzhou 450001, China

Received:2025-03-20 Published:2025-12-04

摘要/Abstract

摘要： 针对高光谱图像分类中面临的小样本问题,本文提出了高光谱图像深度特征提取方法。利用预训练基础模型提取高光谱图像的深度图,作为先验信息与光谱信息进行融合分类。为了充分利用高光谱图像中丰富的光谱信息,采用滑窗法沿着光谱维度提取多个深度图,堆叠后称为深度特征。该方法基于多源遥感图像融合思路,但无需获取精确配准的多源遥感图像,具有即插即用的优点。3个高光谱图像上大量的分类试验验证了该方法的有效性。

关键词: 高光谱图像, 深度特征提取, 小样本分类, 基础大模型

Abstract: A depth feature extraction method for hyperspectral image classification is proposed to address the small sample problem.The proposed method first utilizes a pre-trained base model to extract depth maps from hyperspectral images as prior information,which is then fused with spectral information for classification.To fully exploit the rich spectral information in hyperspectral images,a sliding window approach is employed to extract multiple depth maps along the spectral dimension,which are then stacked to form depth features.The method is based on the concept of multi-source remote sensing image fusion but does not require precisely registered multi-source remote sensing images,offering a plug-and-play advantage.Extensive classification experiments on three hyperspectral image datasets validate the effectiveness of the method.

Key words: hyperspectral image, depth feature extraction, few-shot classification, foundation large model

中图分类号:

P237

秦进春, 裴航, 刘冰, 余岸竹, 陈俊铭, 范俊忆. 融合深度特征的高光谱图像小样本分类方法[J]. 测绘通报, 2025, 0(11): 34-39.

QIN Jinchun, PEI Hang, LIU Bing, YU Anzhu, CHEN Junming, FAN Junyi. Few-shot hyperspectral image classification with depth feature fusion[J]. Bulletin of Surveying and Mapping, 2025, 0(11): 34-39.

参考文献

[1] GONG Zhiqiang,ZHONG Ping,HU Weidong.Statistical loss and analysis for deep learning in hyperspectral image classification[J].IEEE Transactions on Neural Networks and Learning Systems,2021,32(1):322-333.
[2] GHAMISI P,YOKOYA N,LI Jun,et al.Advances in hyperspectral image and signal processing:a comprehensive overview of the state of the art[J].IEEE Geoscience and Remote Sensing Magazine,2017,5(4):37-78.
[3] LIANG Liang,DI Liping,ZHANG Lianpeng,et al.Estimation of crop LAI using hyperspectral vegetation indices and a hybrid inversion method[J].Remote Sensing of Environment,2015,165:123-134.
[4] POUR A B,ZOHEIR B,PRADHAN B,et al.Editorial for the special issue:multispectral and hyperspectral remote sensing data for mineral exploration and environmental monitoring of mined areas[J].Remote Sensing,2021,13(3):519.
[5] SHIMONI M,HAELTERMAN R,PERNEEL C.Hyperspectral imaging for military and security applications:combining myriad processing and sensing techniques[J].IEEE Geoscience and Remote Sensing Magazine,2019,7(2):101-117.
[6] CLARK M L,ROBERTS D A.Species-level differences in hyperspectral metrics among tropical rainforest trees as determined by a tree-based classifier[J].Remote Sensing,2012,4(6):1820-1855.
[7] YUE Shihong,LI Ping,HAO Peiyi.SVM classification:its contents and challenges[J].Applied Mathematics-A Journal of Chinese Universities,2003,18(3):332-342.
[8] RIGATTI S J.Random forest[J].Journal of Insurance Medicine,2017,47(1):31-39.
[9] AHMAD M,KHAN A M,MAZZARA M,et al.A fast and compact 3-D CNN for hyperspectral image classification[J].IEEE Geoscience and Remote Sensing Letters,2022,19:5502205.
[10] BENEDIKTSSON J A,PALMASON J A,SVEINSSON J R.Classification of hyperspectral data from urban areas based on extended morphological profiles[J].IEEE Transactions on Geoscience and Remote Sensing,2005,43(3):480-491.
[11] 高奎亮,刘冰,余岸竹,等.高光谱影像少样例分类的无监督元学习方法[J].测绘学报,2023,52(11):1941-1952.
[12] LI Zhaokui,LIU Ming,CHEN Yushi,et al.Deep cross-domain few-shot learning for hyperspectral image classification[J].IEEE Transactions on Geoscience and Remote Sensing,2022,60:5501618.
[13] XI Bobo,ZHANG Yun,LI Jiaojiao,et al.CTF-SSCL:CNN-transformer for few-shot hyperspectral image classification assisted by semisupervised contrastive learning[J].IEEE Transactions on Geoscience and Remote Sensing,2024,62:5532617.
[14] RANFTL R,LASINGER K,HAFNER D,et al.Towards robust monocular depth estimation:mixing datasets for zero-shot cross-dataset transfer[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2022,44(3):1623-1637.
[15] YANG Lihe,KANG Bingyi,HUANG Zilong,et al.Depth anything:unleashing the power of large-scale unlabeled data[C]//Proceedings of 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Seattle:IEEE,2024:10371-10381.
[16] YANG L,KANG B,HUANG Z,et al.Depth anything v2[J].Advances in Neural Information Processing Systems,2025,37:21875-21911.
[17] RANFTL R,BOCHKOVSKIY A,KOLTUN V.Vision transformers for dense prediction[C]//Proceedings of 2021 IEEE/CVF International Conference on Computer Vision (ICCV).Montrea:IEEE,2022:12159-12168.

融合深度特征的高光谱图像小样本分类方法

Few-shot hyperspectral image classification with depth feature fusion

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 8

编辑推荐

Metrics

本文评价

[1]	许淇, 杨嘉葳, 王继燕. 基于二次平滑和特征加权的高光谱图像分类[J]. 测绘通报, 2025, 0(6): 55-61.
[2]	张佳宝, 谢福鼎, 金翠. 基于超像素图卷积网络的高光谱图像分类[J]. 测绘通报, 2023, 0(1): 84-87.
[3]	徐君, 王彩玲. 结合空间信息选取最优端元组合的混合像元分解[J]. 测绘通报, 2019, 0(7): 6-11.
[4]	宋广钦, 杜正舜, 贺智. 高光谱图像波段选择的改进二进制布谷鸟算法[J]. 测绘通报, 2019, 0(4): 43-48.
[5]	李彩虹, 赵祎霏. 一种高光谱图像的半监督分类方法[J]. 测绘通报, 2018, 0(2): 46-49.
[6]	胡冬翠, 谢福鼎, 杨俊, 张永. 一种基于空间信息和遗传算法的半监督高光谱图像分类方法[J]. 测绘通报, 2018, 0(10): 22-26.
[7]	程志会, 谢福鼎. 基于空间特征与纹理信息的高光谱图像半监督分类[J]. 测绘通报, 2016, 0(12): 56-59,73.
[8]	何晓宁曹建农高坡高美玲. 顶点成分分析法端元提取的改进算法[J]. 测绘通报, 2013, 0(7): 30-34.