留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

融合级联CRFs和U-Net深度学习模型的遥感影像建筑物自动提取

陈嘉浩 邢汉发 陈相龙

陈嘉浩, 邢汉发, 陈相龙. 融合级联CRFs和U-Net深度学习模型的遥感影像建筑物自动提取[J]. 华南师范大学学报(自然科学版), 2022, 54(1): 70-78. doi: 10.6054/j.jscnun.2022011
引用本文: 陈嘉浩, 邢汉发, 陈相龙. 融合级联CRFs和U-Net深度学习模型的遥感影像建筑物自动提取[J]. 华南师范大学学报(自然科学版), 2022, 54(1): 70-78. doi: 10.6054/j.jscnun.2022011
CHEN Jiahao, XING Hanfa, CHEN Xianglong. Automatic Building Extraction from Remote Sensing Images Based on Cascaded CRFs and the U-Net Deep Learning Model[J]. Journal of South China normal University (Natural Science Edition), 2022, 54(1): 70-78. doi: 10.6054/j.jscnun.2022011
Citation: CHEN Jiahao, XING Hanfa, CHEN Xianglong. Automatic Building Extraction from Remote Sensing Images Based on Cascaded CRFs and the U-Net Deep Learning Model[J]. Journal of South China normal University (Natural Science Edition), 2022, 54(1): 70-78. doi: 10.6054/j.jscnun.2022011

融合级联CRFs和U-Net深度学习模型的遥感影像建筑物自动提取

doi: 10.6054/j.jscnun.2022011
基金项目: 

国家自然科学基金项目 41971406

广东省科技计划项目 2018B020207002

详细信息
    通讯作者:

    邢汉发,Email: xinghanfa@163.com

  • 中图分类号: P237

Automatic Building Extraction from Remote Sensing Images Based on Cascaded CRFs and the U-Net Deep Learning Model

  • 摘要: 针对深度学习模型进行建筑物提取时存在的建筑物边缘模糊问题,将级联CRFs(全连接条件随机场)引入到U-Net深度模型中,提出了一种改进的U-Net模型(U-Net+级联CRFs),以用于遥感影像建筑物自动提取:构建级联CRFs并将其引入到U-Net模型的解码层中,从多层结构中学习边界信息,增强模型对建筑物边界的分割能力。并以广东省佛山市为研究区,利用U-Net+级联CRFs、U-Net+CRFs、U-Net、SVM模型进行建筑物提取实验。结果表明U-Net+级联CRFs模型可以有效识别建筑物边界信息,提高建筑物提取的精度:U-Net+级联CRFs模型在准确度、召回率、F1值和交并比4个指标上的均值分别达到了93.1%、87.5%、91.4%和85.1%,均优于U-Net+CRFs、U-Net、SVM模型。
  • 图  1  U-Net网络结构示意图

    Figure  1.  The schematic diagram of U-Net network structure

    图  2  级联CRFs的抽象结构

    Figure  2.  The abstract structure of the cascaded CRFs

    图  3  级联CRFs的U-Net网络结构

    Figure  3.  The U-Net network structure of the cascaded CRFs

    图  4  研究区域位置图

    注:此图基于自然资源部标准地图服务网站的标准地图(审图号:GS(2016)2556号)绘制, 底图无修改。

    Figure  4.  The location map of the study area

    图  5  研究区部分影像建筑物及标签数据集

    Figure  5.  Some image buildings and label data in the study area

    图  6  基于4种模型的建筑物提取可视化对比

    Figure  6.  The visual comparison of building extraction with four models

    图  7  基于4种模型的不同建筑物的提取对比效果

    Figure  7.  The comparison of the effect of extracting different buildings with four models

    图  8  基于4种模型的建筑物边界分割细节

    Figure  8.  The details of building boundary division extracted with four models

    图  9  4种模型的精度评价结果

    Figure  9.  The results of accuracy evaluation of four models

  • [1] 崔卫红, 熊宝玉, 张丽瑶. 多尺度全卷积神经网络建筑物提取[J]. 测绘学报, 2019, 48(5): 597-608.

    CUI W H, XIONG B Y, ZHANG L Y. Multi-scale fully convolutional neural network for building extraction[J]. Acta Geodaetica et Cartographica Sinica, 2019, 48(5): 597-608.
    [2] 林祥国, 张继贤. 面向对象的形态学建筑物指数及其高分辨率遥感影像建筑物提取应用[J]. 测绘学报, 2017, 46(6): 724-733. https://www.cnki.com.cn/Article/CJFDTOTAL-CHXB201706009.htm

    LIN X G, ZHANG J X. Object-based morphological building index for building extraction from high resolution remote sensing imagery[J]. Acta Geodaetica et Cartographica Sinica, 2017, 46(6): 724-733. https://www.cnki.com.cn/Article/CJFDTOTAL-CHXB201706009.htm
    [3] 朱岩彬, 徐启恒, 杨俊涛, 等. 基于全卷积神经网络的高分辨率航空影像建筑物提取方法研究[J]. 地理信息世界, 2020, 27(2): 101-106. https://www.cnki.com.cn/Article/CJFDTOTAL-CHRK202002018.htm

    ZHU Y B, XU Q H, YANG J T, et al. Research on building extraction method of high resolution aerial image based on full convolutional neural network[J]. Geomatics World, 2020, 27(2): 101-106. https://www.cnki.com.cn/Article/CJFDTOTAL-CHRK202002018.htm
    [4] MAGGIORI E, TARABALKA Y, CHARPIAT G, et al. Con-volutional neural networks for large-scale remote sensing image classification[J]. IEEE Transactions on Geoscience & Remote Sensing, 2016, 55(2): 645-657.
    [5] 于浩, 徐汉卿, 李雨佳, 等. 针对无人机影像的农居建筑物快速提取方法初探[J]. 防灾减灾学报, 2020, 36(4): 79-85.

    YU H, XU H Q, LI Y J, et al. Preliminary study on fast extraction method of agricultural residential buildings from UAV images[J]. Seismological Research of Northeast China, 2020, 36(4): 79-85.
    [6] 郝乐川. 高分辨率可见光遥感图像建筑物提取方法研究[D]. 哈尔滨: 哈尔滨工业大学, 2020.

    HAO L C. Research on building extraction from high-reso-lution visible optical remote sensing images[D]. Harbin: Harbin Institute of Technology, 2020.
    [7] 张士诚, 李新萍, 盛奇. 结合方向梯度和支持向量机的立面窗户提取方法[J]. 测绘科学, 2020, 45(9): 159-164. https://www.cnki.com.cn/Article/CJFDTOTAL-CHKD202009025.htm

    ZHANG S C, LI X P, SHENG Q. Elevation window extraction method combining directional gradient and support vector machine[J]. Science of Surveying and Ma-pping, 2020, 45(9): 159-164. https://www.cnki.com.cn/Article/CJFDTOTAL-CHKD202009025.htm
    [8] 陈元鹏, 罗明, 彭军还, 等. 基于网格搜索随机森林算法的工矿复垦区土地利用分类[J]. 农业工程学报, 2017, 33(14): 250-257. doi: 10.11975/j.issn.1002-6819.2017.14.034

    CHEN Y P, LUO M, PENG J H, et al. Land use classification of industrial and mining reclamation area based on grid search random forest algorithm[J]. Transactions of the Chinese Society of Agricultural Engineering, 2017, 33(14): 250-257. doi: 10.11975/j.issn.1002-6819.2017.14.034
    [9] 朱海洲, 贾银山. 基于支持向量机的遥感图像分类研究[J]. 科学技术与工程, 2010, 10(15): 3659-3663. doi: 10.3969/j.issn.1671-1815.2010.15.023

    ZHU H Z, JIA Y S. Remote sensing image classification based on Support Vector Machines[J]. Science Technology and Engineering, 2010, 10(15): 3659-3663. doi: 10.3969/j.issn.1671-1815.2010.15.023
    [10] TARI G, JESSEN L, KENNELLY P, et al. Surface ma-pping of the Milh Kharwah salt diapir to better understand the subsurface petroleum system in the Sab'atayn Basin, onshore Yemen[J]. Arabian Journal of Geosciences, 2018, 11(15): 428-438. doi: 10.1007/s12517-018-3785-y
    [11] HUERTAS A, NEVATIA R. Detecting buildings in aerial images[J]. Computer Vision, Graphics, and Image Processing, 1988, 41(2): 131-152. doi: 10.1016/0734-189X(88)90016-3
    [12] 韩东成. 基于高分二号遥感卫星的建筑物信息提取研究[D]. 合肥: 中国科学技术大学, 2019.

    HAN D C. Research on building Information Extraction based on Gaofen-2 Remote Sensing satellite[D]. Hefei: University of Science and Technology of China, 2019.
    [13] MCGLONE J C, SHUFELT J A. Projective and object space geometry for monocular building extraction[C]//Procee-dings of IEEE Computer Society Conference on Computer Vision & Pattern Recognition. Seattle, WA: IEEE, 2002.
    [14] LHOMME S, DONG-CHEN H E, WEBER C, et al. A new approach to building identification from very high spatial-resolution images[J]. International Journal of Remote Sensing, 2009, 30(5/6): 1341-1354.
    [15] LECUN Y, BOTTOU L. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324. doi: 10.1109/5.726791
    [16] 张顺, 龚怡宏, 王进军. 深度卷积神经网络的发展及其在计算机视觉领域的应用[J]. 计算机学报, 2019, 42(3): 453-482. https://www.cnki.com.cn/Article/CJFDTOTAL-JSJX201903001.htm

    ZHANG S, GONG Y H, WANG J J, et al. Development of deep convolution neural network and its application in the filed of computer vision[J]. Computer Journal, 2019, 42(3): 453-482. https://www.cnki.com.cn/Article/CJFDTOTAL-JSJX201903001.htm
    [17] GUO Z L, SHAO X W, XU Y W, et al. Identification of village building via Google Earth images and supervised machine learning methods[J]. Remote Sensing, 2016, 8(4): 271/1-15. doi: 10.3390/rs8040271
    [18] LONG J, SHELHAMER E, DARRELL T. Fully convolutional networks for semantic segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 39(4): 640-651.
    [19] RONNEBERGER O, FISCHER P, BROX T. U-Net: con-volutional networks for biomedical image segmentation[C]//Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015. Cham, Switzerland: Springer, 2015: 234-241.
    [20] 朱辉, 秦品乐. 基于多尺度特征结构的U-Net肺结节检测算法[J]. 计算机工程, 2019, 45(4): 254-261.

    ZHU H, QIN P L. U-Net pulmonary nodule detection algorithm based on multi-scale feature structure[J]. Computer Engineering, 2019, 45(4): 254-261.
    [21] 刘浩, 骆剑承, 黄波, 等. 基于特征压缩激活Unet网络的建筑物提取[J]. 地球信息科学学报, 2019, 21(11): 1779-1789. doi: 10.12082/dqxxkx.2019.190285

    LIU H, LUO J C, HUANG B, et al. Building extraction based on SE-Unet[J]. Journal of Geo-information Science, 2019, 21(11): 1779-1789. doi: 10.12082/dqxxkx.2019.190285
    [22] 伍广明, 陈奇, SHIBASAKI R, 等. 基于U型卷积神经网络的航空影像建筑物检测[J]. 测绘学报, 2018, 47(6): 864-872.

    WU G M, CHEN Q, SHIBASAKI R, et al. High precision building detection from aerial imagery using a U-Net like convolutional architecture[J]. Acta Geodaetica et Cartographica Sinica, 2018, 47(6): 864-872.
    [23] 张浩然, 赵江洪, 张晓光. 利用U-net网络的高分遥感影像建筑提取方法[J]. 遥感信息, 2020, 35(3): 143-150. https://www.cnki.com.cn/Article/CJFDTOTAL-YGXX202003020.htm

    ZHANG H R, ZHAO J H, ZHANG X G. High-resolution image building extraction using U-net neural network[J]. Remote Sensing Information, 2020, 35(3): 143-150. https://www.cnki.com.cn/Article/CJFDTOTAL-YGXX202003020.htm
    [24] 任欣磊, 王阳萍, 杨景玉, 等. 基于改进U-net的遥感影像建筑物提取[J]. 激光与光电子学进展, 2019, 56(22): 195-202.

    REN X L, WANG Y P, YANG J Y, et al. Building detection from remote sensing images based on improved U-net[J]. Laser and Optoelectronics Progress, 2019, 56(22): 195-202.
    [25] 马永杰, 李雪燕, 宋晓凤. 基于改进深度卷积神经网络的交通标志识别[J]. 激光与光电子学进展, 2018, 55(12): 250-257. https://www.cnki.com.cn/Article/CJFDTOTAL-JGDJ201812024.htm

    MA Y J, LI X Y, SONG X F. Traffic sign recognition based on improved deep convolutional neural network[J]. Laser and Optoelectronics Progress, 2018, 55(12): 250-257. https://www.cnki.com.cn/Article/CJFDTOTAL-JGDJ201812024.htm
  • 加载中
图(9)
计量
  • 文章访问数:  251
  • HTML全文浏览量:  115
  • PDF下载量:  81
  • 被引次数: 0
出版历程
  • 收稿日期:  2021-03-17
  • 刊出日期:  2022-02-25

目录

    /

    返回文章
    返回