| 290 | 0 | 170 |
| 下载次数 | 被引频次 | 阅读次数 |
针对遥感影像水边线提取任务，基于U-Net、PSP-Net和DeepLabV3+三种常用的语义分割模型，结合三波段(RGB)和四波段(RGB+近红外波段)遥感影像数据，通过平均像素准确率(Mean Pixel Accuracy, MPA)、交并比(Intersection over Union, IoU)和平均IoU(mean IoU,MIoU)指标评估模型性能，系统对比了不同模型以及波段数量对水体提取精度的影响。结果表明，DeepLabV3+的整体提取效果最佳，MPA达到90.32%,加入近红外波段后，模型提取精度显著提高，U-Net和DeepLabV3+的mIoU分别提升3.92%和3.0%,有效改善了水体提取过程中模型漏提和错提的现象。
Abstract：In this study, the performance of three widely-used semantic segmentation models(U-Net, PSP-Net, and DeepLabV3+) is systematically evaluated for waterline extraction tasks using both three-band(RGB) and four-band(RGB+near-infrared band) remote sensing imagery. The effects of different model architectures and spectral band combinations on extraction accuracy are compared through Mean Pixel Accuracy(MPA), Intersection over Union(IoU), and mean IoU(MIoU) metrics. Experimental results demonstrate that DeepLabV3+ achieves optimal extraction performance with an MPA of 90.32%. The incorporation of near-infrared bands significantly enhances model accuracy, improving mIoU by 3.92% for U-Net and 3.0% for DeepLabV3+ respectively. This spectral augmentation effectively mitigates under-extraction and mis-extraction phenomena during water body identification processes.
[1] SMALL C,NICHOLLS R J.A Global Analysis of Human Settlement in Coastal Zones[J].Journal of Coastal Research,2003,19(3):584-599.
[2] ZHANG K Q,DOUGLAS B C,LEATHERMAN S P.Global Warming and Coastal Erosion[J].Climatic Change,2004,64(1):41-58.
[3] YASIR M,HUI S,HUANG B H,et al.Coastline Extraction and Land Use Change Analysis Using Remote Sensing (RS) and Geographic Information System (GIS) Technology—A Review of the Literature[J].Reviews on Environmental Health,2020,35(4):453-460.
[4] 马小峰，赵冬至，邢小罡，等.海岸线卫星遥感提取方法研究[J].海洋环境科学，2007(2):185-189.
[5] LI R,DI K C,MA R J.A Comparative Study of Shoreline Mapping Techniques[M].[S.l.]:CRC Press,2004.
[6] BOAK E H,TURNER I L.Shoreline Definition and Detection:A Review[J].Journal of Coastal Research,2005,21(4):688-703.
[7] DING X S,SHAN X J,CHEN Y L,et al.Dynamics of Shoreline and Land Reclamation from 1985 to 2015 in the Bohai Sea,China[J].Journal of Geographical Sciences,2019,29(12):2031-2046.
[8] MCFEETERS S K.The Use of the Normalized Difference Water Index (NDWI) in the Delineation of Open Water Features[J].International Journal of Remote Sensing,1996,17(7):1425-1432.
[9] 刘苏宜，黄对，王建军，等.基于NDWI及扩展指数的地表水分变化监测——以宿鸭湖水库为例[J].水利水运工程学报，2024(6):103-114.
[10] 宋昊，潘鑫，谢文英，等.基于修正的增强归一化差异水体指数的富营养化水体信息提取研究[J].湖泊科学，2025,37(2):378-389.
[11] 谭玉敏，槐建柱，唐中实.一种融合边缘信息的面向对象遥感图像分割方法[J].光谱学与光谱分析，2010,30(6):1624-1627.
[12] KAUR B,GARG A.Mathematical Morphological Edge Detection for Remote Sensing Images.International Conference on Electronics Computer Technology[C]//2011 3rd International Conference on Electronics Computer Technology.Kanyakumari:IEEE,2011:324-327.
[13] 郑云柯，王俊霖，张世涛.基于改进Canny算子的水域边界自动提取方法研究[J].软件导刊，2023,22(9):167-173.
[14] GOH G B,HODAS N O,VISHNU A.Deep Learning for Computational Chemistry[J].Journal of Computational Chemistry,2017,38(16):1291-1307.
[15] XIE S N,YANG T B,WANG X Y,et al.Hyper-class Augmented and Regularized Deep Learning for Fine-grained Image Classification[C]//2015 IEEE Conference on Computer Vision & Pattern Recognition.Boston:IEEE,2015:2645-2654.
[16] SUN Y,WANG X,TANG X.Deep Learning Face Representation from Predicting 10,000 Classes[C]//2014 IEEE Conference on Computer Vision and Pattern Recognition.Columbus:IEEE,2014:1891-1898.
[17] LITJENS G,KOOI T,BEJNORDI B E,et al.A Survey on Deep Learning in Medical Image Analysis[J].Medical Image Analysis,2017,42(9):60-88.
[18] POPLIN R,VARADARAJAN A V,BLUMER K,et al.Predicting Cardiovascular Risk Factors from Retinal Fundus Photographs using Deep Learning[EB/OL].(2017-08-31)[2025-01-11].https://arxiv.org/abs/1708.09843.
[19] 司浚豪，邵峰晶，隋毅，等.基于深度学习的遥感图像水边线提取方法与应用[J].海洋环境科学，2022,41(2):309-315.
[20] 田萱，王亮，丁琪.基于深度学习的图像语义分割方法综述[J].软件学报，2019,30(2):440-468.
[21] 龚蒙，张永宏，孙书林，等.DPL-Net:用于高分辨率遥感图像的水体提取网络[J/OL].计算机工程与应用，1-11[2025-01-25].http://kns.cnki.net/kcms/detail/11.2127.TP.20241025.1430.004.html.
[22] 王婷婷，程熙，李斌，等.基于高分辨率遥感影像的城市水体语义分割和制图——以中国成都为例[J/OL].物探化探计算技术，1-9[2025-01-25].http://kns.cnki.net/kcms/detail/51.1242.P.20240628.1722.002.html.
[23] 白倩，罗小波，母仕林.基于TA-UNet3+的高分辨率遥感图像地表水体提取[J/OL].计算机工程与应用，1-13[2025-01-25].http://kns.cnki.net/kcms/detail/11.2127.TP.20240909.1703.024.html.
[24] 项文成，王洲，王建平，等.基于假设检验的SAR图像水域分割增强方法[J].无线电工程，2024,54(12):2895-2901.
[25] RONNEBERGER O,FISCHER P,BROX T.U-Net:Convolutional Networks for Biomedical Image Segmentation[EB/OL].(2015-05-18)[2025-01-11].https://arxiv.org/abs/1505.04597.
[26] SALAMA W M,ALY M H.Deep Learning in Mammography Images Segmentation and Classification:Automated CNN Approach[J].AEJ-Alexandria Engineering Journal,2021,60(5):4701-4709.
[27] LI W,RAJ A N J,TJAHJADI T,et al.Digital Hair Removal by Deep Learning for Skin Lesion Segmentation[J].Pattern Recognition,2021,117:107994.
[28] 郑海洋，于淼，于晓鹏.VCDG-UNet模型在遥感图像分割中的应用[J].无线电工程，2025,55(1):94-104.
[29] ZHAO H S,SHI J P,QI X J,et al.Pyramid Scene Parsing Network[EB/OL].(2016-12-04)[2025-01-11].https://arxiv.org/abs/1612.01105.
[30] CHEN L C,ZHU Y K,PAPANDREOU G,et al.Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation[C]//European Conference on Computer Vision.Munich:Spinger,2018:833-851.
[31] 东方星.我国高分卫星与应用简析[J].卫星应用，2015(3):44-48.
[32] 孙伟富，马毅，张杰，等.不同类型海岸线遥感解译标志建立和提取方法研究[J].测绘通报，2011(3):41-44.
[33] MITCHELL T M.Machine Learning[M].[S.l.]:McGraw-Hill Education,1997.
基本信息:
中图分类号:P237;P332;TP751
引用信息:
[1]田森,蔺楠.基于语义分割模型与遥感影像波段扩展的水边线提取[J].无线电工程,2025,55(06):1256-1264.
基金信息:
陕西省秦创原“科学家+工程师”队伍建设项目(2024QCY-KXJ-093)~~
2025-03-21
2025-03-21
2025-03-21