融合多模态数据的河道遥感地物识别

doi:10.13474/j.cnki.11-2246.2025.1231

测绘通报 ›› 2025, Vol. 0 ›› Issue (12): 178-183.doi: 10.13474/j.cnki.11-2246.2025.1231

• 测绘地理信息技术应用案例 • 上一篇

融合多模态数据的河道遥感地物识别

王超¹, 付强¹, 崔志芳¹, 唐甜²

1. 长江水利委员会水文局长江中游水文水资源勘测局, 湖北武汉 430010;
2. 武汉天宝耐特科技有限公司, 湖北武汉 430070

收稿日期:2025-07-22 发布日期:2025-12-31
通讯作者: 崔志芳。E-mail:cui@whu.edu.cn
作者简介:王超(1989—),男,高级工程师,主要从事河道测绘与无人机遥感。E-mail:45586077@qq.com
基金资助:
国家重点研发计划(2023YFC3209502);长江水利委员会水文局科技创新基金(SWJ-25CJX24)

Remote sensing feature identification of river channel by fusing multimodal data

WANG Chao¹, FU Qiang¹, CUI Zhifang¹, TANG Tian²

1. Middle Changjiang River Bureau of Hydrology and Water Resources Survey, Bureau of Hydrology of Changjiang Water Resource Commission, Wuhan 430010, China;
2. Wuhan Trimble Net Technology Co., Ltd., Wuhan 430070, China

Received:2025-07-22 Published:2025-12-31

摘要/Abstract

摘要： 针对无人机遥感中河道地物因光谱混淆与阴影遮挡导致的地物分类精度不足的问题,本文提出了一种基于DeepLabV3+的RGB-DSM多模态数据融合语义分割方法。该方法通过整合无人机RGB影像的光谱信息与激光点云生成的DSM高程信息,引入“光谱-结构”特征协同机制,实现对水体、堤坡、树木等河道典型地物的像素级精准识别。试验结果表明,与RGB相比,RGB-DSM多模态数据输入下,模型像素精度(PA)由93.47%提升至95.06%,平均交并比(mIoU)由81.60%提高至84.38%,其中树木类别IoU提升显著。视觉对比显示,DSM高程信息的加入有效改善了道路边界模糊、树木阴影误分类等问题,显著增强了复杂地物的空间结构表达能力。本文证明了RGB-DSM多模态融合在提升河道地物语义分割精度方面具有独特优势。

关键词: 无人机遥感, 河道地物识别, 语义分割, DeepLabV3+, RGB-DSM

Abstract: Aiming at the lack of feature classification accuracy of river features in UAV remote sensing due to spectral confusion and shadow masking,this study proposes a semantic segmentation method based on DeepLabV3+ for RGB-DSM multimodal data fusion.The method integrates the spectral information of RGB images from UAVs and the DSM elevation information generated from laser point clouds,and introduces the “spectral-structural” feature synergy mechanism to realize the pixel-level accurate recognition of typical river features,such as water bodies,embankment slopes,trees,and so on.The experimental results show that,compared with RGB,the pixel accuracy (PA)of the model is increased from 93.47% to 95.06%,and the mean intersection ratio (mIoU)is increased from 81.60% to 84.38%,with a significant increase in the IoU of the tree category under the RGB-DSM multimodal data input.The visual comparison shows that the inclusion of DSM elevation information effectively improves the problems of road boundary blurring and tree shadow misclassification,and significantly enhances the spatial structure representation of complex features.This paper demonstrates that RGB-DSM multimodal fusion has unique advantages in improving the semantic segmentation accuracy of river features.

Key words: UAV remote sensing, river feature recognition, semantic segmentation, DeepLabV3+, RGB-DSM

中图分类号:

P237

王超, 付强, 崔志芳, 唐甜. 融合多模态数据的河道遥感地物识别[J]. 测绘通报, 2025, 0(12): 178-183.

WANG Chao, FU Qiang, CUI Zhifang, TANG Tian. Remote sensing feature identification of river channel by fusing multimodal data[J]. Bulletin of Surveying and Mapping, 2025, 0(12): 178-183.

参考文献

[1] WOODGET A S, AUSTRUMS R, MADDOCK I P, et al.Drones and digital photogrammetry: from classifications to continuums for monitoring river habitat and hydromorphology[J].WIREs Water, 2017, 4(4): e1222.
[2] 黄育华, 吴念辉, 陈杰, 等.基于无人机全景影像的河道岸线地物变化检测方法[J].人民长江, 2025, 56(3): 230-236.
[3] CHAO Zhenhua, FANG Xuan, NA Jiaming, et al.A collaborative sensing system for farmland water conservancy project maintenance through integrating satellite, aerial, and ground observations[J].Water, 2021, 13(16): 2163.
[4] KUCHARCZYK M, HUGENHOLTZ C H.Remote sensing of natural hazard-related disasters with small drones: global trends, biases, and research opportunities[J].Remote Sensing of Environment, 2021, 264:112577.
[5] MOHD DAUD S M S, MOHD YUSOF M Y P, HEO C C, et al.Applications of drone in disaster management: a scoping review[J].Science ＆ Justice, 2022, 62(1):30-42.
[6] 许强, 郭晨, 董秀军.地质灾害航空遥感技术应用现状及展望[J].测绘学报, 2022, 51(10): 2020-2033.
[7] 肖斌, 罗浩, 张恒宾, 等.多尺度融合卷积的轻量化Transformer无人机地物识别模型[J].郑州大学学报(理学版), 2024, 56(1): 32-39.
[8] LI Ziran, NAMIKI A, SUZUKI S, et al.Application of low-altitude UAV remote sensing image object detection based on improved YOLOv5[J].Applied Sciences, 2022, 12(16): 8314.
[9] CAO Zhen, KOOISTRA L, WANG Wensheng, et al.Real-time object detection based on UAV remote sensing: a systematic literature review[J].Drones, 2023, 7(10): 620.
[10] 雷相达, 王宏涛, 赵宗泽.整合迁移学习与全卷积网络的小样本机载激光雷达点云分类[J].中国激光, 2021, 48(16):138-149.
[11] 黄远程, 陈领, 江宇, 等.融合注意力和多尺度表达的机载激光点云分类[J].测绘科学, 2022, 47(11): 137-144.
[12] SHEYKHMOUSA M, MAHDIANPARI M, GHANBARI H, et al.Support vector machine versus random forest for remote sensing image classification: a meta-analysis and systematic review[J].IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2020, 13: 6308-6325.
[13] AKAR Ö, GVNGÖR O.Classification of multispectral images using random forest algorithm[J].Journal of Geodesy and Geoinformation, 2012, 1(2): 105-112.
[14] CHARBUTY B, ABDULAZEEZ A.Classification based on decision tree algorithm for machine learning[J].Journal of Applied Science and Technology Trends, 2021, 2(1): 20-28.
[15] BANDYOPADHYAY S, MAULIK U, MUKHOPADHYAY A.Multiobjective genetic clustering for pixel classification in remote sensing imagery[J].IEEE Transactions on Geoscience and Remote Sensing, 2007, 45(5): 1506-1511.
[16] RONNEBERGER O, FISCHER P, BROX T.U-Net: convolutional networks for biomedical image segmentation[M]//Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015.Cham: Springer International Publishing, 2015: 234-241.
[17] BADRINARAYANAN V, KENDALL A, CIPOLLA R.SegNet: a deep convolutional encoder-decoder architecture for image segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(12): 2481-2495.
[18] ZHAO Jinling, LI Zheng, LEI Yu, et al.Application of UAV RGB images and improved PSPNet network to the identification of wheat lodging areas[J].Agronomy, 2023, 13(5): 1309.
[19] CHEN L, PAPANDREOU G, KOKKINOS I, et al.Semantic image segmentation with deep convolutional nets and fully connected CRFs[EB/OL].[2025-05-22].https://arxiv.org/abs/1412.7062.
[20] CHEN L C, PAPANDREOU G, KOKKINOS I, et al.DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834-848.
[21] CHEN L, PAPANDREOU G, SCHROFF F, et al.Rethinking atrous convolution for semantic image segmentation[EB/OL].[2025-05-22].https://arxiv.org/abs/1706.05587.
[22] CHEN L C, ZHU Yukun, PAPANDREOU G, et al.Encoder-decoder with atrous separable convolution for semantic image segmentation[C]//Proceedings of 2018 European Conference on Computer Vision (ECCV).Cham: Springer, 2018: 833-851.
[23] SHELHAMER E, LONG J, DARRELL T.Fully convolutional networks for semantic segmentation[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition.[S.l.]:IEEE, 2017:640-651.

融合多模态数据的河道遥感地物识别

Remote sensing feature identification of river channel by fusing multimodal data

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	邓小龙, 黄志海, 郭波. 改进分割大模型的桥梁缆索损伤语义分割方法[J]. 测绘通报, 2025, 0(9): 112-117.
[2]	王靖凯, 葛星彤, 李兆博, 丁翔, 彭玲. 基于改进HRNet的高速公路路域内光伏板信息提取[J]. 测绘通报, 2025, 0(5): 74-78,99.
[3]	李建, 王健, 王雷, 李敏, 杨立克, 赵艺龙. 双重注意力机制的电力走廊点云语义分割[J]. 测绘通报, 2025, 0(4): 127-133.
[4]	马于博, 张爱国, 王浩宇, 刘帅琪, 靳镜宇, 沈占锋, 李均力. 基于U-ConvHDNet模型的戈壁砾幕层提取[J]. 测绘通报, 2025, 0(4): 68-74.
[5]	张开洲, 马瑞峰, 贾鑫. 基于语义分割网络和特征匹配的复杂山区高速公路路面线性要素智能提取方法[J]. 测绘通报, 2025, 0(4): 164-169.
[6]	符强, 钟振, 纪元法, 任风华. 基于动态场景的实时语义SLAM算法[J]. 测绘通报, 2025, 0(4): 27-33.
[7]	苟长龙, 庞敏, 杨扬. 改进的U-Net卷积网络在遥感影像地物分类中的应用[J]. 测绘通报, 2025, 0(3): 150-155.
[8]	赵兴旺, 赵妍, 刘超, 刘春阳. 基于改进的DeepLabV3+网络的Sentinel-1影像水体提取[J]. 测绘通报, 2025, 0(3): 66-70.
[9]	陈忠超, 孙俊英, 谭登澳, 李娟, 成其换, 杨铭珂. 基于视觉基础模型优化的U-Net++遥感语义分割方法[J]. 测绘通报, 2025, 0(12): 121-125,162.
[10]	赵效祖, 苟长龙, 杨扬. 基于CBAM增强的轻量级遥感影像语义分割方法[J]. 测绘通报, 2025, 0(10): 36-42.
[11]	张伟, 张朝龙, 王本林, 蔡安宁. 基于深度学习的多尺度无人机遥感图像道路提取[J]. 测绘通报, 2024, 0(6): 77-81.
[12]	王鸿雁, 车向红, 徐辛超, 徐胜华, 李洪胜. 利用DeepLabv3+模型提取分析街景图像绿视率——以北京三环内为例[J]. 测绘通报, 2024, 0(3): 88-94.
[13]	黄应华, 董振川, 李昊, 陈壮, 刘长睿, 张献州. 城市竣工测绘典型要素语义分割PointNet++深度学习模型适用性分析[J]. 测绘通报, 2024, 0(2): 85-89.
[14]	黄金彩, 李诗逸, 石岩. 利用行车记录仪视频提取路面车道线[J]. 测绘通报, 2024, 0(12): 1-5.
[15]	李锋, 薛梅, 詹勇, 杨元. 基于分割一切模型SAM的实景三维场景语义分割[J]. 测绘通报, 2024, 0(12): 101-105.