基于随机补片和DeepLabV3+的建筑物遥感图像变化检测

doi:10.13278/j.cnki.jjuese.20200159

吉林大学学报(地球科学版) ›› 2021, Vol. 51 ›› Issue (6): 1932-1938.doi: 10.13278/j.cnki.jjuese.20200159

• 地球探测与信息技术 • 上一篇

基于随机补片和DeepLabV3+的建筑物遥感图像变化检测

王民水^1,2, 孔祥明³, 陈学业^1,4, 杨国东^1,2, 王明常^1,2, 张海明²

1. 自然资源部城市国土资源监测与仿真重点实验室, 广东深圳 518000;
2. 吉林大学地球探测科学与技术学院, 长春 130026;
3. 山东省物化探勘查院, 济南 250013;
4. 深圳市数字城市工程研究中心, 广东深圳 518034

收稿日期:2020-07-08 出版日期:2021-11-26 发布日期:2021-11-24
通讯作者: 王明常(1975-),男,教授,博士,主要从事遥感与地理信息系统研究,E-mail:wangmc@jlu.edu.cn E-mail:wangmc@jlu.edu.cn
作者简介:王民水(1989-),男,工程师,硕士,主要从事遥感图像处理和实验教学工作,E-mail:1543383519@qq.com
基金资助:
自然资源部城市国土资源监测与仿真重点实验室开放基金项目（KF-2019-04-080）；自然资源部地面沉降监测与防治重点实验室开放基金项目（KLLSMP201901）；吉林省教育厅"十三五"科学研究规划项目（JJKH20200999KJ）；国家自然科学基金项目（42171407）

Remote Sensing Image Change Detection Based on Random Patches and DeepLabV3+ Network

Wang Minshui^1,2, Kong Xiangming³, Chen Xueye^1,4, Yang Guodong^1,2, Wang Mingchang^1,2, Zhang Haiming²

1. Key Laboratory of Urban Land Resources Monitoring and Simulation, Ministry of Natural Resources, Shenzhen 518000, Guangdong, China;
2. College of GeoExploration Science and Technology, Jilin University, Changchun 130026, China;
3. Shandong Institute of Geophysical & Geochemical Exploration, Jinan 250013, China;
4. Shenzhen Research Center of Digital City Engineering, Shenzhen 518034, Guangdong, China

Received:2020-07-08 Online:2021-11-26 Published:2021-11-24
Supported by:
Supported by the Open Fund of Key Laboratory of Urban Land Resources Monitoring and Simulation, Ministry of Natural Resources (KF-2019-04-080), the Key Laboratory of Land Subsidence Monitoring and Prevention, Ministry of Natural Resources (KLLSMP201901), the Scientific Research Project of the 13th Five-Year Plan of Jilin Province Education Department (JJKH20200999KJ) and the National Natural Science Foundation of China(42171407)

摘要/Abstract

摘要： 为有效解决传统遥感图像变化检测预处理复杂的问题，提出一种基于随机补片和DeepLabV3+的建筑物遥感图像变化检测方法。以ResNet50特征提取网络为基础，创建DeepLabV3+语义分割网络，并在图像和标签中创建大小为224像素×224像素的随机补片作为网络输入，训练建筑物提取网络；修改建筑物提取网络输入层为6通道，通过矩阵运算将两期遥感图像转换为一幅6通道非RGB图像，利用转换后的非RGB图像进行网络训练并验证变化检测精度。实验1利用ENVI5.3软件，采用马氏距离法进行变化检测；实验2采用改进的U-Net网络和随机补片，完成网络训练和精度验证；实验3使用实验2的训练数据和验证数据，采用随机补片和DeepLabV3+网络进行变化检测网络训练及精度验证。实验结果表明，该方法实验1、实验2、实验3建筑物变化检测平均交并比分别为24.43%、83.14%、89.90%，边界轮廓匹配分数分别为61.47%，80.24%、96.51%。

关键词: 随机补片, DeepLabV3+网络, 语义分割, 建筑物变化检测

Abstract: In order to effectively preprocess the traditional remote sensing image change detection, we proposed a change detection method of building remote sensing image based on random patches and DeeplabV3+. This method builds a DeepLabV3+ semantic segmentation network based on the ResNet50, which is a feature extraction network, crops the random patches of 224 pixels×224 pixels in the image and label them as the network input to train the building extraction network,and then, modify the input layer of the building extraction network to six channels. The two-phase remote sensing images are converted into a 6-channel non-RGB image through matrix operation, which are used for network training and validating the change detection accuracy. In Experiment 1, the Mahalanobis distance classification method was used to detect the change by ENVI5.3 software. In Experiment 2, the improved U-Net network and random patches were used to complete the network training and accuracy verification. Experiment 3 used the training data and verification data of Experiment 2, and used random patches and DeepLabV3+ network to train the change detection network and verify the accuracy.The results of Experiment 1, 2, and 3 show that the average intersection-over-union of this method is 24.43%, 83.14%, and 89.90% respectively, and the boundary matching score is 61.47%, 80.24%, and 96.51% respectively.

Key words: random patches, DeepLabV3+network, semantic segmentation, building change detection

中图分类号:

TP79

王民水, 孔祥明, 陈学业, 杨国东, 王明常, 张海明. 基于随机补片和DeepLabV3+的建筑物遥感图像变化检测[J]. 吉林大学学报(地球科学版), 2021, 51(6): 1932-1938.

Wang Minshui, Kong Xiangming, Chen Xueye, Yang Guodong, Wang Mingchang, Zhang Haiming. Remote Sensing Image Change Detection Based on Random Patches and DeepLabV3+ Network[J]. Journal of Jilin University(Earth Science Edition), 2021, 51(6): 1932-1938.

参考文献

[1] 李亮,舒宁,王凯,等. 融合多特征的遥感影像变化检测方法[J]. 测绘学报,2014, 43(9):945-953, 959. Li Liang, Shu Ning, Wang Kai, et al. Change Detection Method for Remote Sensing Images Based on Multi-Features Fusion[J]. Acta Geodaetica Cartographica Sinica, 2014, 43(9):945-953, 959.
[2] 眭海刚,冯文卿,李文卓,等.多时相遥感影像变化检测方法综述[J].武汉大学学报(信息科学版),2018,43(12):1885-1898. Sui Haigang, Feng Wenqing, Li Wenzhuo, et al. Review of Change Detection Methods for Multi-Temporal Remote Sensing Imagery[J]. Geomatics and Information Science of Wuhan University, 2018, 43(12):1885-1898.
[3] 王明常,牛雪峰,杨毅恒,等.长白山地区景观格局过程模拟预测研究[J].吉林大学学报(地球科学版),2009, 39(5):947-952. Wang Mingchang, Niu Xuefeng, Yang Yiheng, et al. Simulation and Predicted Research on Changbai Mountain Landscape Pattern Process[J]. Journal of Jilin University (Earth Science Edition), 2009, 39(5):947-952.
[4] 王明常,郭鑫,王凤艳,等.基于FLUS的长春市土地利用动态变化与预测分析[J].吉林大学学报(地球科学版),2019, 39(6):1795-1804. Wang Mingchang, Guo Xin, Wang Fengyan, et al. Dynamic Change and Predictive Analysis Changchun City Based on FLUS Model[J]. Journal of Jilin University (Earth Science Edition), 2019, 39(6):1795-1804.
[5] 范荣双,陈洋,徐启恒,等.基于深度学习的高分辨率遥感影像建筑物提取方法[J].测绘学报,2019, 48(1):34-41. Fan Rongshuang, Chen Yang, Xu Qiheng, et al. A High-Resolution Remote Sensing Image Building Extraction Method Based on Deep Learning[J]. Acta Geodaetica Cartographica Sinica, 2019, 48(1):34-41.
[6] Kemker R, Salvaggio C, Kanan C. Algorithms for Semantic Segmentation of Multispectral Remote Sensing Imagery Using Deep Learning[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2018, 145(11):60-77.
[7] Zhao J, Gong M, Liu J, et al. Deep Learning to Classify Difference Image for Image Change Detection[C]//International Joint Conference on Neural Networks.[S. l.]:IEEE, 2014:411-417.
[8] Zhan Y, Fu K, Yan M, et al. Change Detection Based on Deep Siamese Convolutional Network for Optical Aerial Images[J]. IEEE Geoence and Remote Sensing Letters, 2017, 14(10):1845-1849.
[9] 袁立,袁吉收,张德政.基于DeepLab-V3+的遥感影像分类[J].激光与光电子学进展,2019,56(15):236-243. Yuan Li, Yuan Jishou, Zhang Dezheng. Remote Sensing Image Classification Based on DeepLab-V3+[J]. Laser & Optoelectronics Progress, 2019, 56(15):236-243.
[10] 朱春宇,王明常,王凤艳,等.基于深度置信网络与数学形态学融合的遥感影像建筑物变化检测[J].科学技术与工程,2020, 20(8):3157-3163. Zhu Chunyu, Wang Mingchang, Wang Fengyan, et al. Building Change Detection Based on Deep Belief Networks and Mathematical Morphology Fusion[J].Science Technology and Engineering, 2020, 20(8):3157-3163.
[11] Wang Mingchang, Zhang Haiming, Sun Weiwei, et al. A Coarse-to-Fine Deep Learning Based Land Use Change Detection Method for High-Resolution Remote Sensing Images[J]. Remote Sensing, 2020, 12(12):1933.
[12] He K, Zhang X, Ren S, et al. Deep Residual Learning for Image Recognition[C]//IEEE Conference on Computer Vision & Pattern Recognition.[S. l.]:IEEE Computer Society, 2016:770-778.
[13] Long J, Shelhamer E, Darrell T. Fully Convolutional Networks for Semantic Segmentation[J]. Transactions on Pattern Analysis & Machine Intelligence, 2014, 39(4):640-651.
[14] Yang L, Wang H, Yan K. Building Extraction of Multi-Source Data Based on Deep Learning[C]//2019 IEEE 4th International Conference on Image. Xiamen:Vision and Computing (ICIVC), 2019:296-300.
[15] Emmanuel Maggiori, Yuliya Tarabalka, Guillaume Charpiat, et al. "Can Semantic Labeling Methods Generalize to Any City? The Inria Aerial Image Labeling Benchmark"[C]//IEEE International Geoscience and Remote Sensing Symposium. Fort Worth:IEEE International Symposium on Geoscience and Remote Sensing (IGARSS), 2017:3226-3229.
[16] Ji Shunping, Wei Shiqing, Lu Meng. Fully Convolutional Networks for Multi-Source Building Extraction from an Open Aerial and Satellite Imagery Data Set[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(1):574-586.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于随机补片和DeepLabV3+的建筑物遥感图像变化检测

Remote Sensing Image Change Detection Based on Random Patches and DeepLabV3+ Network

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

Metrics

本文评价

推荐阅读 10

[1]	牛莹, 赵欣怡, 周云轩, 田波, 王利花. 基于MODIS的长江口表层水体盐度时空分异[J]. 吉林大学学报(地球科学版), 2019, 49(5): 1486-1495.
[2]	李刚, 孙桂华, 姚永坚, 朱博勤, 张耀明. 三沙湾海岸线时空演变[J]. 吉林大学学报(地球科学版), 2019, 49(1): 196-205.
[3]	何梦颖, 梅西, 张训华, 刘健, 郭兴伟, 郑洪波. 南黄海陆架区CSDP-1孔沉积物碎屑锆石U-Pb年龄物源判别[J]. 吉林大学学报(地球科学版), 2019, 49(1): 85-95.
[4]	熊凯, 宫兆宁, 张磊, 赵文吉. 再生水补水条件下土壤全氮空间分布特征[J]. 吉林大学学报(地球科学版), 2017, 47(6): 1829-1837.
[5]	陈圣波, 于亚凤, 杨金中, 王楠, 梦华. 基于实测光谱指数法的ASTER遥感数据岩性信息提取[J]. 吉林大学学报(地球科学版), 2016, 46(3): 938-944.
[6]	黄绍霖，徐涵秋，王琳. CPF变化对Landsat TM/ETM+辐射校正结果的影响[J]. 吉林大学学报(地球科学版), 2014, 44(4): 1382-1387.
[7]	王明常，牛雪峰，陈圣波，王亚楠，汪自军. 基于DART模型的PROBA/CHRIS数据叶面积指数反演[J]. 吉林大学学报(地球科学版), 2013, 43(3): 1033-1039.
[8]	崔瀚文，姜琦刚，邢宇，徐驰，林楠. 32 a来气候扰动下中国沙质荒漠化动态变化[J]. 吉林大学学报(地球科学版), 2013, 43(2): 582-591.
[9]	陈勇，何中发，黎兵，赵宝成. 崇明东滩潮沟发育特征及其影响因素定量分析[J]. 吉林大学学报(地球科学版), 2013, 43(1): 212-219.
[10]	陈圣波，刘彦丽，杨倩，周超，赵靓. 植被覆盖区卫星高光谱遥感岩性分类[J]. 吉林大学学报(地球科学版), 2012, 42(6): 1959-1965.
[11]	杨佳佳, 姜琦刚, 赵静, 徐言, 孟翔冲. 基于改进的SVM技术和高光谱遥感的标准矿物定量计算[J]. J4, 2012, 42(3): 864-871.
[12]	陈永良, 李学斌, 林楠. 遥感图像像素级异常识别的一种方法[J]. J4, 2012, 42(3): 881-886.
[13]	王利花, 周云轩, 田波. 基于TM和ETM+影像数据的东沙环礁珊瑚礁监测[J]. J4, 2011, 41(5): 1630-1637.
[14]	贺金鑫, Jonathan Li, 闫浩文. 一种面向水质监测的地球空间传感器网络架构[J]. J4, 2011, 41(4): 1262-1266.
[15]	李华朋, 张树清, 孙妍, 刘春悦. 集成多时相ETM+影像的证据推理湿地遥感分类[J]. J4, 2011, 41(4): 1246-1252.