Please wait a minute...
img

官方微信

遥感技术与应用  2021, Vol. 36 Issue (2): 285-292    DOI: 10.11873/j.issn.1004-0323.2021.2.0285
CNN 专栏     
基于U-Net的高分辨率遥感图像土地利用信息提取
陈妮(),应丰,王静,李健
中国电建集团华东勘测设计研究院有限公司,浙江 杭州 311122
Research on Land Use Information Extraction based on U-Net
Ni Chen(),Feng Ying,Jing Wang,Jian Li
Powerchina Huadong Engineering Corporation Limited,Hangzhou 311122,China
 全文: PDF(18949 KB)   HTML
摘要:

随着现代遥感技术的迅速发展,遥感图像的质量和数量得到了显著的提升,新技术带来的高分辨率遥感图像所蕴含的信息也更加丰富,如何利用人工智能手段辅助挖掘这些丰富的信息也成为了遥感图像分析与理解的重要内容。与此同时,以深度卷积神经网络为代表的人工智能技术在图像处理领域大放异彩。得益于类人眼的分层卷积池化模型,深度卷积神经网络可以在图像分割和分类等任务上取得优异的结果。因此采用U-Net为代表的深度卷积神经网络对2 m的高分辨率遥感影像进行了特征提取、分割和分类,不同于传统基于手工设定图像特征的方法,U-Net可以自动对海量高分辨率的遥感图像进行特征提取,从而充分挖掘高分辨率遥感影像中复杂的非线性特征、光谱特征和纹理特征。实验结果表明:利用训练好的U-Net模型对新昌县土地利用分类计算时间为55.7 s,分类准确率可达90.95%,Kappa系数为0.86。U-Net模型可以快速、精确地提取高分辨率遥感影像中的地表覆盖特征,得到高精度的土地利用分类结果,说明将该模型应用于遥感影像土地利用分类提取有着广阔前景。

关键词: 全卷积神经网络U?Net土地利用分类高分辨遥感图像    
Abstract:

With the rapid development of modern remote sensing technology, remote sensing image with high quality and quantity has been significantly promoted, new technology of high resolution remote sensing images contain more abundant information, how to make full use of the means of artificial intelligence auxiliary to mine these abundant information has become one of the important researches in remote sensing image analysis and understanding. At the same time, represented by deep convolutional neural networks based Artificial Intelligence (AI) technology is brilliant in the field of image processing. Thanks to the layer-wised convolutional and pooling structures which mimces human brain retinal systems, deep convolutional neural network can achieve excellent performance in image segmentation and classification. So this paper proposed a U-Net based model to extract features from high resolution remote sensing images with 2 m spatial resolution. Different from traditional methods based on hand craft image features, the proposed model can be automatically applied on massive amounts of high resolution remote sensing image feature extraction, it can also exert complicated nonlinear characteristics of high resolution remote sensing image with the help of the spectral features and texture features. The experimental results show that the time of using the U-Net model to calculate the land use classification of Xinchang County is 55.7s, and the accuracy is 90.95%, and the kappa coefficient is 0.86. U-Net model can quickly and accurately obtain the land cover features in high-resolution remote sensing images, and can get high-precision land use classification results, which shows that the deep learning into remote sensing image land use classification extraction has a broad prospect.

Key words: Full convolutional neural network    U-Net    Land use classification    High resolution remote sensing images
收稿日期: 2019-12-17 出版日期: 2021-05-24
ZTFLH:  TP79  
作者简介: 陈妮(1988-),女,浙江诸暨人,博士,工程师,主要从事水土保持与遥感应用研究。E?mail:chen_n3@hdec.com
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章  
陈妮
应丰
王静
李健

引用本文:

陈妮,应丰,王静,李健. 基于U-Net的高分辨率遥感图像土地利用信息提取[J]. 遥感技术与应用, 2021, 36(2): 285-292.

Ni Chen,Feng Ying,Jing Wang,Jian Li. Research on Land Use Information Extraction based on U-Net. Remote Sensing Technology and Application, 2021, 36(2): 285-292.

链接本文:

http://www.rsta.ac.cn/CN/10.11873/j.issn.1004-0323.2021.2.0285        http://www.rsta.ac.cn/CN/Y2021/V36/I2/285

图1  本研究的U-Net的网络结构
图2  基于U-Net的土地利用分类流程图
图 3  本文获取的土地利用高分辨率遥感样本示例集
图4  训练准确率和误差变化
图5  土地利用分类精度对比

地类U-NetSegNet
IoU(%)Precision(%)Recall(%)IoU(%)Precision(%)Recall(%)
1耕地91.9893.3398.4590.9893.0497.63
2园地88.8292.0596.287.7991.3595.74
3林地94.9696.1398.7394.6595.9298.62
4草地87.1489.9696.5286.6989.6996.28
5建设用地94.8496.5398.1994.3396.8697.31
6交通运输用地87.6789.4597.7887.0489.0197.52
7水域89.0292.6795.7689.2393.3195.33
8其他土地88.1591.9795.586.2390.8594.43
9扰动用地90.5392.6997.4990.2792.5197.39
总体精度90.9593.297.4290.3192.9596.95
Kappa系数0.860.84
表1  土地利用分类结果的精度评估
1 Chen J, Chen J, Liao A P, et a1. Global Land Cover Mapping at 30 m Resolution: A POK-based Operational Approach [J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2015, 103: 7-27.
2 Zhang Can, Xu Hanqiu, Zhang Hao, et a1. Fractional Vegetation Cover Change and Its Ecological Effect Assessment in a Typical Reddish Soil Region of Southeastern China: Changting County, Fujian Province [J]. Journal of Natural Resources, 2015, 30(6): 917-928.
2 张灿, 徐涵秋, 张好, 等. 南方红壤典型水土流失区植被覆盖度变化及其生态效应评估——以福建省长汀县为例[J]. 自然资源学报, 2015, 30(6): 917-928.
3 Du Xianyuan, Chen Hongkun, Weng Yibin, et a1. Application of High-resolution Remote Sensing Image to Monitoring Soil Erosion in Shale Gas Development [J]. Natural Gas Industry, 2019, 39(12): 161-167.
3 杜显元, 陈宏坤, 翁艺斌, 等. 高分遥感影像在页岩气开发水土流失监测中的应用[J]. 天然气工业, 2019, 39(12): 161-167.
4 Ma Liping, Cheng Hui. The Application of High-resolution Remote Sensing and UAV Technology in Soil and Water Conservation Supervision [J]. Construction & Design for Project, 2019(7): 117-118.
4 马利苹, 程辉. 水土保持监管中的高分遥感以及无人机技术应用[J]. 工程建设与设计, 2019(7): 117-118.
5 Feng Yuyun. Application of Modern Surveying and Mapping Technology in Agricultural Land Management [J]. Value Engineering, 2020, 39(6): 249-250.
5 丰玉云. 现代测绘技术在农业土地管理的应用分析[J]. 价值工程, 2020, 39(6): 249-250.
6 Li Chaojun. Dynamic Analysis of Agricultural Land Use in Nanjing based on Remote Sensing Image [J]. Journal of Shanxi Agricultural Sciences, 2016, 44(12): 1896-1898,1902.
6 李超君. 基于遥感图像的南京市农业用地动态分析[J]. 山西农业科学, 2016, 44(12): 1896-1898,1902.
7 Li Chungan, Chen Qi, Tan Bizeng. Large-scale Forest Resources Monitoring by Means of Spatial Sampling of Hi-resolution Remote Sensing Images [J]. Forest Resources Management, 2009(2): 106-110,127.
7 李春干, 陈琦, 谭必增. 基于卫星遥感数据空中抽样的大尺度森林资源动态监测 [J]. 林业资源管理, 2009(2): 106-110,127.
8 Lu Xiaohui. Forest Resource Estimation based on Remote Sensing Technology in Suizhong County of Liaoning Province[J]. Inner Mongolia Forestry Investigation and Design, 2019, 42(5): 60-61,91,94.
8 陆小辉. 基于遥感技术的辽宁省绥中县森林资源估算[J]. 内蒙古林业调查设计, 2019, 42(5): 60-61,91,94.
9 Qu Ran, Wang Changzuo, Liu Huiming, et a1. Myanmar and China in Areas Bordering Deforestation Remote Sensing Monitoring[J]. Environment and Sustainable Development, 2012, 37(5): 98-102.
9 屈冉, 王昌佐, 刘慧明, 等.缅甸与我国接壤地区森林砍伐遥感监测分析 [J]. 环境与可持续发展, 2012, 37(5): 98-102.
10 Zhu Yun, Ling Zhigang, Zhang Yuqiang. Research Progress and Prospect of Machine Vision Technology [J]. Journal of Graphics, 2020, 41(6): 871-890.
10 朱云, 凌志刚, 张雨强. 机器视觉技术研究进展及展望 [J]. 图学学报, 2020, 41(6): 871-890.
11 Xu Hui, Zhu Yuhua, Zhen Tong, et a1. Survey of Image Semantic Segmentation Methods based on Deep Neural Network [J]. Journal of Frontiers of Computer Science and Technology, 2021, 15(1): 47-59.
11 徐辉, 祝玉华, 甄彤, 等. 深度神经网络图像语义分割方法综述[J]. 计算机科学与探索, 2021, 15(1) :47-59.
12 Shi Caijuan, Zhang Weiming, Chen Houru, et a1. Survey of Salient Object Detection based on Deep Learning [J]. Journal of Frontiers of Computer Science and Technology, 2021, 15(2): 219-232.
12 史彩娟, 张卫明, 陈厚儒, 等. 基于深度学习的显著性目标检测综述[J]. 计算机科学与探索, 2021, 15(2): 219-232.
13 Lecun Y, Bottou L, Bengio Y, et al. Gradient-based Learning Applied to Document Recognition [J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.
14 Russakovsky O, Deng J, Su H, et al. ImageNet Large Scale Visual Recognition Challenge[J]. International Journal of Computer Vision, 2015, 115(3): 211-252.
15 Vakalopoulou M, Karantzalos K, Komodakis N, et al. Building Detection in very High Resolution Multispectral Data with Deep Learning Features[C]∥ IGARSS 2015-2015 IEEE International Geoscience and Remote Sensing Symposium. IEEE, 2015: 1873-1876.
16 Guo Z, Shao X, Xu Y, et a1. Identification of Village Building via Google Earth Images and Supervised Machine Learning Methods[J]. Remote Sensing,2016,8(4):271. doi:10. 3390/rs8040271.
doi: 10. 3390/rs8040271
17 Feng Jiawen, Zhang Limin, Deng Xiangyang. Image Segmentation based on Multi-source Fusion FCN [J]. Application Research of Computers, 2018, 35(9): 323-326.
17 冯家文, 张立民, 邓向阳. 基于多源融合FCN的图像分割[J]. 计算机应用研究, 2018, 35(9): 323-326.
18 Jonathan L, Evan S, Darrell T. Fully Convolutional Networks for Semantic Segmentation[C]∥ Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Boston, MA:IEEE, 2015: 3431-3440.
19 Ronneberger O, Fischer P, Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation[C]∥ In: International Conference on Medical Image Computing and Computer-Assisted Intervention, Springer, Berlin, 2005:234-241.
20 Zhang Z, Liu Q, Wang Y. Road Extraction by Deep Eesidual U-Net [J]. IEEE Geoscience and Remote Sensing Letters, 2018, 15(5): 749-753.
21 Li R, Liu W, Yang L, et al. DeepUNet: A Deep Fully Convolutional Network for Pixel-level Sea-land Segmentation [J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018, 11(11): 3954-3962.
22 Xu Y, Wu L, Xie Z, et al. Building Extraction in very High Resolution Remote Sensing Imagery Using Deep Learning and Guided Filters[J]. Remote Sensing, 2018, 10(1): 144. doi:10.3390/rs10010144.
doi: 10.3390/rs10010144
23 Wang Junqiang, Li Jiansheng, Ding Bo, et al. Performance Analysis of Semantic Segmentation Method based on Deep Learning in Remote Sensing Image Segmentation [J]. Computer Measurement & Control, 2019, 27(7): 231-235.
23 王俊强, 李建胜, 丁波, 等. 深度学习语义分割方法在遥感影像分割中的性能分析 [J]. 计算机测量与控制, 2019, 27(7): 231-235.
24 Yang Jianyu, Zhou Zhenxu, Du Zhenrong, et al. Rural Construction Land Extraction from High Spatial Resolution Remote Sensing Image based on SegNet Semantic Segmentation Model [J]. Transactions of the Chinese Society of Agricultural Engineering, 2019, 35(5): 251-258.
24 杨建宇, 周振旭, 杜贞容, 等. 基于SegNet语义模型的高分辨率遥感影像农村建设用地提取 [J]. 农业工程学报, 2019, 35(5): 251-258.
25 Deng Tianmin, Wang Lin, Yang Qizhi, et al. Lane Line Detection Method based on Improved SegNet Algorithm [J]. Science Technology and Engineering, 2020, 20(36): 14988-14993.
25 邓天民, 王琳, 杨其芝, 等. 基于改进SegNet算法的车道线检测方法 [J]. 科学技术与工程, 2020, 20(36): 14988-14993.
[1] 程雨婷,刘昭华,鹿琳琳,刘士彪,李庆亭. 一带一路沿海超大城市热岛时空特征遥感分析[J]. 遥感技术与应用, 2020, 35(5): 1197-1205.
[2] 张峰极,吴艳兰,姚雪东,梁泽毓. 基于改进DenseNet网络的多源遥感影像露天开采区智能提取方法[J]. 遥感技术与应用, 2020, 35(3): 673-684.
[3] 宋晓阳,姜小三,江东,黄耀欢,万华伟,王昌佐. 基于面向对象的高分影像分类研究[J]. 遥感技术与应用, 2015, 30(1): 99-105.
[4] 李 娜,闫冬梅,张 丽. 天津滨海新区—曹妃甸典型地区土地利用演变分析[J]. 遥感技术与应用, 2012, 27(2): 289-295.
[5] 蒋李兵,王 壮,胡卫东. 一种基于可变夹角链码的靠岸舰船目标检测方法[J]. 遥感技术与应用, 2007, 22(1): 88-94.
[6] 傅文杰, 洪金益 , 林明森. 基于光谱相似尺度的支持向量机遥感土地利用分类[J]. 遥感技术与应用, 2006, 21(1): 25-30.
[7] 徐素兰 , 李四海 . 航空像片数字图像土地利用分类和制图[J]. 遥感技术与应用, 1996, 11(1): 54-61.