基于深度学习的生产建设项目扰动图斑自动识别分类

doi:10.16843/j.sswc.2022.06.014

摘要
图/表
参考文献
相关文章 (15)

全文: HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要扰动图斑是生产建设项目水土保持信息化监管基础数据。针对扰动图斑传统人机交互目视解译效率低、成果不统一等问题,基于深度学习原理,构建生产建设项目扰动图斑自动识别分类卷积神经网络模型,用以提高扰动图斑解译生产效率和成果质量。然后确定深度学习模型关键超参数－优化器算法、学习速率和批大小最优值。在此基础上经过150个训练轮次得到生产建设项目扰动图斑自动识别分类卷积神经网络模型,模型综合性能评价指标－精度值和损失值分别为0.9526和0.1670。模型在"检验样本集"应用效果表明:模型识别分类总体精度为97.52%,扰动样本查准率和查全率分别为72.44%和83.90%;模型识别分类结果与真实情况基本一致,漏分类、误分类比例相对较低,具有较强的泛化能力。这说明深度学习模型用于生产建设项目扰动图斑自动识别分类是实际可行的。研究成果为扰动图斑解译生产提供一种新方法,可为生产建设项目水土保持信息化监管提供重要技术支撑。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	金平伟
	黄俊
	姜学兵
	亢庆
	杨胜权
	林丽萍
	杨平
	罗志铖
	李乐
	寇馨月
	刘斌

关键词 ：卷积神经网络, 深度学习模型, 扰动图斑, 高分遥感影像, 水土保持, 监管

Abstract：[Background] The supervision and management of soil erosion caused by construction projects is an important legal responsibility and social management function of the water administrative department. Satellite remote sensing imagery is an important method. As the basic data of supervision and management work, the disturbance patch boundary data of construction projects is currently mainly obtained by manual visual interpretation, which has low work efficiency and large cost investment. [Methods] Based on the deep learning principle, a convolutional neural network was constructed to realize automatic identification of disturbance zone, so as to achieve the goal of automatic production of disturbance patch boundary data of construction projects. The remote sensing image of China's high-resolution No. 1 satellite with a resolution of 2 meters was used in this study. The constructed convolutional neural network contained 13 convolutional layers, 5 max-pooling layers, 1 global average pooling layer, 3 fully connected layers and 2 dropout layers. The activation functions in the middle and classification layer were RELU and Softmax, respectively. The training and testing samples were 5131 and 22923, and the proportion of positive samples in the training and testing sample sets were 15.38% and 4.69%, respectively. [Results] The results showed that the optimizer, learning rate and batch size have a significant impact on the model training accuracy. The optimizer and learning rate have a negligible impact on the training time, and the batch size has a significant impact on it. In this study, the Adagrad optimizer, a 10^-4 learning rate, and a 16 batch size were the optimal choices to obtain the best trained model with accuracy and loss of 0.9526 and 0.1670, respectively. The testing sample is used to validate the model, and the results showed, the average overall accuracy of the model is 97.52%, the average precision and recall of positive samples (classified samples of interest) are 72.44% and 83.90%, respectively, and the average F₁ score is 77.73%. In general, the model recognition and classification results are basically consistent with the actual situation. [Conclusions] This study provides a new method for the production of disturbance patch boundary data of construction projects, which greatly improves work efficiency, reduces input costs, and enhances the efficiency of soil and water conservation supervision and management. In the follow-up, the production of model training sample data should be strengthened, and the model parameters should be further revised to improve the accuracy of model application.

Key words： convolutional neural network deep learning model disturbed patch high-resolution remote sensing image soil and water conservation supervision and management

收稿日期: 2021-04-14

PACS:	TP183
	P407.8

基金资助:广东省水利科技创新项目“CSLE方程B因子多时空尺度研究及其在广东省水土流失动态监测中的应用”(2020-25);贵州省水利科技创新项目“遥感大数据和AI智能识别技术在贵州省水土保持监管中的示范应用”(KJ202004);珠江水利科学研究院科技创新自立项目“基于深度学习的生产建设项目扰动图斑自动识别提取技术研究与应用”(2021ky021)

通讯作者: 黄俊(1983-),男,博士,高级工程师。主要研究方向:水土保持与遥感。E-mail:hjnwsauf@qq.com E-mail: hjnwsauf@qq.com

作者简介: 金平伟(1976-),男,硕士,高级工程师。主要研究方向:水土保持。E-mail:hale.888@163.com

引用本文:

金平伟, 黄俊, 姜学兵, 亢庆, 杨胜权, 林丽萍, 杨平, 罗志铖, 李乐, 寇馨月, 刘斌. 基于深度学习的生产建设项目扰动图斑自动识别分类[J]. 中国水土保持科学, 2022, 20(6): 116-125.
JIN Pingwei, HUANG Jun, JIANG Xuebing, KANG Qing, YANG Shengquan, LIN Liping, YANG Ping, LUO Zhicheng, LI Le, KOU Xinyue, LIU Bin. Automatic recognition and classification of construction projects' disturbed patches based on deep learning. SSWC, 2022, 20(6): 116-125.

链接本文:

http://journal12.magtechjournal.com/Jweb_stbc/CN/10.16843/j.sswc.2022.06.014 或 http://journal12.magtechjournal.com/Jweb_stbc/CN/Y2022/V20/I6/116

[1]	闫佳杰, 史明昌, 高志强. 生产建设项目水土保持监管空间信息系统研究[J]. 中国农业大学学报, 2018, 23(1): 143. YAN Jiajie, SHI Mingchang, GAO Zhiqiang. Study on spatial information system fro soil and water conservation supervision in production and construction projects[J]. Journal of China Agricultural University, 2018, 23(1): 143.
[2]	姜德文. 水土保持强监管目标任务及方法探讨[J]. 中国水利, 2019, 1(1): 13. JIANG Dewen. Discussion on the objectives, tasks and methods of strong supervision on soil and water conservation[J]. China Water Resources, 2019, 1(1): 13.
[3]	亢庆, 姜德文, 赵院, 等. 生产建设项目水土保持"天地一体化"动态监管关键技术体系[J]. 中国水土保持, 2016(11): 4. KANG Qing, JIANG Dewen, ZHAO Yuan, et al. Dynamic supervision key technology system of soil conservation in the construction projects based on the "air-ground integration"[J]. Soil and Water Conservation in China, 2016(11): 4.
[4]	赵俊侠, 任婧宇, 杨亚娟, 等. 生产建设项目水土保持"天地一体化"监管存在问题及建议[J]. 中国水土保持, 2019(6): 53. ZHAO Junxia, REN Jingyu, YANG Yajuan, et al. Questions and suggestions of soil and water conservation supervision and management technology in construction projects based on the "air-ground integration"[J].Soil and Water Conservation in China, 2019(6): 53.
[5]	曹林林, 李海涛, 韩颜顺, 等. 卷积神经网络在高分遥感影像分类中的应用[J]. 测绘科学, 2016, 41(9): 170. CAO Linlin, LI Haitao, HAN Yanshun, et al. Application of convolutional neural networks in classification of high resolutionn remote sensing imagery[J]. Science of Surveying and Mapping, 2016, 41(9): 170.
[6]	胡杰, 张莹, 谢仕义. 国产遥感影像分类技术应用研究进展综述[J]. 计算机工程与应用, 2021, 57(3): 1. HU Jie, ZHANG Ying, XIE Shiyi. Summary of research progress on application of domestic remote sensing image-classification technology[J]. Computer Engineering and Applications, 2021, 57(3): 1.
[7]	宋廷强, 李继旭, 张信耶. 基于深度学习的高分辨率遥感图像建筑物识别[J]. 计算机工程与应用, 2020, 56(8): 26. SONG Tingqiang, LI Jixu, ZHANG Xinye. Building recognition in high-resolution remote sensing image based on deep learning[J]. Computer Engineering and Applications, 2020, 56(8): 26.
[8]	SIRINUKUNWATTANA K, RAZA S E A, TSANG Y, et al. Locality sensitive deep learning for detection and classification of nuclei in routine colon cancer histology images[J]. IEEE Transactions on Medical Imaging, 2016(35): 1196.
[9]	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. FAN Rongshuang, CHEN Yang, XU Qiheng, et al. A high resolution remote sensing image building extraction method based on deep leanring[J]. Acta Geodaetica et Cartographica Sinica, 2019, 48(1): 34.
[10]	王鑫, 李可, 宁晨, 等. 基于深度卷积神经网络和多核学习的遥感图像分类方法[J]. 电子与信息学报, 2019, 41(5): 1098. WANG Xin, LI Ke, NING Chen, et al. Remote sensing image classification method based on deep convolution neural network and multi-kernel learning[J]. Journal of Electronics & Information Technology, 2019, 41(5): 1098.
[11]	CHEN Yang, FAN Rongshuang, BILAL Muhammad, et al. Multilevel cloud detection for high-resolution remote sensing imagery using multiple convolutional neural networks[J]. ISPRS International Journal of Geo-Information, 2018, 7(5):1.
[12]	裴亮, 刘阳, 谭海, 等. 基于改进的全卷积神经网络的资源三号遥感影像云检测[J]. 激光与光电子学进展, 2019, 56(5): 226. PEI Liang, LIU Yang, TAN Hai, et al. Cloud detection of ZY3 satellite remote sensing images based on improved fully convolutional neural networks[J]. Laser & Optoelectronics Progress, 2019, 56(5): 226.
[13]	YANG X, CHEN Z, LI B, et al. A fast and precise method for large-scale land-use mapping based on deep learning[C]. IGARSS 2019-2019 IEEE International Geoscience and Remote Sensing Symposium,2019: 5913.
[14]	马永建, 汪传建, 赵庆展, 等.基于GF-1遥感影像的荒漠区耕地分类与提取方法[J].石河子大学学报(自然科学版), 2021: 1[DOI:10.13880/j.cnki.65-1174/n.2021.21.012]. MA Yongjian, WANG Chuanjian, ZHAO Qingzhan, et al. Study on extraction of cultivated land in desert area with GF-1 remote sensing image based on U-Net model[J]. Journal of Shihezi University (Natural Science), 2021: 1[DOI:10.13880/j.cnki.65-1174/n.2021.21.012].
[15]	陈周, 费鲜芸, 高祥伟, 等.高分辨率遥感影像分割的城市绿地提取研究[J]. 测绘通报, 2020(12):17. CHEN Zhou, FEI Xianyun, GAO Xiangwei, et al. Extraction of urban green space with high resolution remote sensing image segmentation[J]. Bulletin of Surveying and Mapping, 2020(12): 17.
[16]	胡乃勋,陈涛,甄娜,等. 基于卷积神经网络的面向对象露天采场提取[J].遥感技术与应用, 2021,36(2):265. HU Naixun,CHEN Tao,ZHEN Na,et al. Object-oriented open pit extraction based on convolutional neural network[J]. Remote Sensing Technology and Application, 2021,36(2):265.
[17]	李小占, 马本学, 喻国威, 等. 基于深度学习与图像处理的哈密瓜表面缺陷检测[J]. 农业工程学报, 2021, 37(1): 223. LI Xiaozhan, MA Benxue, YU Guowei, et al. Surface defect detection of Hami melon using deep learning and image processing[J]. Transactions of the CSAE, 2021, 37(1): 223.
[18]	钟亮, 郭熙, 国佳欣, 等. 基于不同卷积神经网络模型的红壤有机质高光谱估算[J]. 农业工程学报, 2021, 37(1): 203. ZHONG Liang, GUO Xi, GUO Jiaxin, et al. Hyperspectral estimation of organic matter in red soil using different convolutional neural network models[J]. Transactions of the CSAE, 2021, 37(1):203.
[19]	李就好, 林乐坚, 田凯, 等. 改进Faster R-Cnn的田间苦瓜叶部病害检测[J]. 农业工程学报, 2020, 36(12): 179. LI Jiuhao, LIN Lejian, TIAN Kai, et al. Detection of leaf diseases of balsam pear in the field based on improved Faster R-CNN[J]. Transactions of the CSAE, 2020, 36(12): 179.
[20]	任守纲, 贾馥玮, 顾兴健, 等. 反卷积引导的番茄叶部病害识别及病斑分割模型[J]. 农业工程学报, 2020, 36(12): 186. REN Shougang, JIA Fuwei, GU Xingjian, et al. Recognition and segmentation model of tomato leaf diseases based on deconvolution-guiding[J]. Transactions of the CSAE, 2020, 36(12): 186.
[21]	谢为俊, 魏硕, 郑招辉, 等. 采用深度多分支模型融合网络的胡萝卜缺陷识别与分割[J]. 农业工程学报, 2021, 37(2): 177. XIE Weijun, WEI Shuo, ZHENG Zhaohui, et al. Classification and segmentation of defect carrots using deep multi-branch models fusion network[J]. Transactions of the CSAE, 2021, 37(2):177.
[22]	HINTON G E, SRIVASTAVA N, KRIZHEVSKY A, et al. Improving neural networks by preventing co-adaptation of feature detectors[J]. 2012: p. arXiv:1207.0580.
[23]	DUCHI J, HAZAN E, SINGER Y. Adaptive subgradient methods for online learning and stochastic optimization[J]. Journal of Machine Learning Research, 2011(12): 2121.
[24]	KINGMA D, BA J. Adam: A method for stochastic optimization[R]. International Conference on Learning Representations, 2014.