|本期目录/Table of Contents|

[1]付 饶,何 敬,刘 刚.基于改进YOLOv4的2021年海地7.2级地震震后滑坡识别[J].地震研究,2023,46(02):300-307.[doi:10.20015/j.cnki.ISSN1000-0666.2023.0012]
　FU Rao,HE Jing,LIU Gang.Landslide Recognition After the 2021 Haiti MS7.2 Earthquake Based on the Improved YOLOv4 Algorithm[J].Journal of Seismological Research,2023,46(02):300-307.[doi:10.20015/j.cnki.ISSN1000-0666.2023.0012]
点击复制

基于改进YOLOv4的2021年海地7.2级地震震后滑坡识别(PDF/HTML)

《地震研究》[ISSN:1000-0666/CN:53-1062/P]

卷:
46
期数:
2023年02期
页码:
300-307
栏目:
出版日期:
2023-06-01

文章信息/Info

Title:
Landslide Recognition After the 2021 Haiti MS7.2 Earthquake Based on the Improved YOLOv4 Algorithm
作者:
付 饶1; 何 敬1; 刘 刚1; 2
(1.成都理工大学 地球科学学院,四川 成都 610059; 2.成都理工大学 地质灾害防治与地质环境保护国家重点实验室,四川 成都 610059)
Author(s):
FU Rao1; HE Jing1; LIU Gang1; 2
(1.School of Earth Sciences,Chengdu University of Technology,Chengdu 610059,Sichuan,China)(2.State Key Laboratory of Geological Disaster Prevention and Geological Environment Protection,Chengdu University of Technology,Chengdu 610059,Sichuan,China)
关键词:
YOLOv4;  海地地震;  滑坡识别;  高分影像
Keywords:
the YOLOv4 algorithm;  the Haiti Earthquake;  landslide identification;  high resolution image
分类号:
P642.22
DOI:
10.20015/j.cnki.ISSN1000-0666.2023.0012
摘要:
以国产高分二号影像为数据源,利用改进的YOLOv4算法对2021年海地7.2级地震诱发的滑坡进行识别。为提升模型的识别效率,用MobileNetv3替换了YOLOv4的骨干网络CSPDarknet53,并用深度可分离卷积替代YOLOv4中的普通卷积,优化了模型参数和网络结构。结果表明:改进后的YOLOv4算法目标识别精度达到91.37%,比普通YOLOv4检测速度提高了6.19 f/s,精度提高了5.24%,模型参数大小减少了80%。改进后的方法对滑坡的检测精度高于原YOLOv4算法,得到的滑坡位置更为准确,具有轻量化和实时性更高的优势,可为应急救援和灾情评估提供更加可靠的数据。
Abstract:
Rapid identification of seismic landslides is important for emergency rescue and loss assessment.On August 14th,2021,a 7.2-magnitude earthquake occurred in Haiti,inducing a large number of landslides.In this paper,the improved YOLOv4 algorithm is used to identify the landslides induced by the Haiti MS7.2 Earthquake using the domestically produced high-fraction 2 images as the data source.To improve the recognition efficiency of the model,the backbone network CSPDarknet53 of Yolov4 is replaced with MobileNetv3,and the ordinary convolution in the YOLOv4 is replaced with depth-separable convolution to optimize the model parameters and network structure.The improved YOLOv4 algorithm achieves 91.37% of the accuracy of target recognition,6.19 f/s(5.24%)higher than the detection speed of the normal YOLOv4,providing more reliable data for emergency rescue and disaster assessment.

参考文献/References:


巨袁臻,许强,金时超,等.2020.使用深度学习方法实现黄土滑坡自动识别[J].武汉大学学报(信息科学版),45(11):1747-1755.
林齐根,邹振华,祝瑛琦,等.2017.基于光谱、空间和形态特征的面向对象滑坡识别[J].遥感技术与应用,32(5):931-937.
刘旭霞,秦芮.2021.7.2级地震令海地“摇摇欲坠”[N].环球时报,2021-08-16(2).
牛全福,程维明,兰恒星,等.2010.玉树地震滑坡灾害的遥感提取与分布特征分析[C]//中国灾害防御协会.全国突发性地质灾害应急处置与灾害防治技术高级研讨会论文集.北京:中国灾害防御协会,7.
许冲,戴福初,陈剑,等.2009.汶川MS8.0地震重灾区次生地质灾害遥感精细解译[J].遥感学报,13(4):754-762.
Benjdira B,Khursheed T,Koubaa A, et al.2019.Car detection using unmanned aerial vehicles:Comparison between faster r-cnn and yolov3[C]//2019 1st International Conference on Unmanned Vehicle Systems-Oman(UVS).IEEE,1-6.
Bialas J,Oommen T,Rebbapragada U, et al.2016.Object-based classification of earthquake damage from high-resolution optical imagery using machine learning[J].Journal of Applied Remote Sensing,10(3):036025.
Bochkovskiy A,Wang C Y,Liao H Y M.2020.Yolov4:Optimal speed and accuracy of object detection[J].arXiv preprint arXiv:2004.10934.
Cheng L,Li J,Duan P, et al.2021a.A small attentional YOLO model for landslide detection from satellite remote sensing images[J].Landslides,18(8):2751-2765.
Cheng Z,Gong W,Tang H, et al.2021b.UAV photogrammetry-based remote sensing and preliminary assessment of the behavior of a landslide in Guizhou,China[J].Engineering Geology,289:106172.
Everingham M,Van Gool L,Williams C K I, et al.2010.The pascal visual object classes(voc)challenge[J].International journal of computer vision,88(2):303-338.
He K,Zhang X,Ren S, et al.2015.Spatial pyramid pooling in deep convolutional networks for visual recognition[J].IEEE transactions on pattern analysis and machine intelligence,37(9):1904-1916.
Howard A G,Zhu M,Chen B, et al.2017.Mobilenets:Efficient convolutional neural networks for mobile vision applications[J].arXiv preprint arXiv,doi:10.48550/arXiv.1704.04861.
Howard A,Sandler M,Chu G, et al.2019.Searching for MobileNetv3[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision,1314-1324.
Huang Y,Zhao L.2018.Review on landslide susceptibility mapping using support vector machines[J].Catena,165:520-529.
Kavzoglu T,Colkesen I,Sahin E K.2019.Machine learning techniques in landslide susceptibility mapping:a survey and a case study[J].Landslides:Theory,practice and modelling,283-301,doi:10.10071978-3-319-77377-3-13.
Keyport R N,Oommen T,Martha T R, et al.2018.A comparative analysis of pixel-and object-based detection of landslides from very high-resolution images[J].International journal of applied earth observation and geoinformation,64:1-11.
Lin T Y,Goyal P,Girshick R, et al.2017.Focal loss for dense object detection[C]//Proceedings of the IEEE international conference on computer vision,2980-2988.
Liu W,Anguelov D,Erhan D, et al.2016.Ssd:Single shot multibox detector[C]//European conference on computer vision.Springer,Cham,21-37.
Micheletti N,Foresti L,Robert S, et al.2014.Machine learning feature selection methods for landslide susceptibility mapping[J].Mathematical geosciences,46(1):33-57.
Redmon J,Divvala S,Girshick R, et al.2016.You only look once:Unified,real-time object detection[C]//Proceedings of the IEEE conference on computer vision and pattern recognition,779-788.
Redmon J,Farhadi A.2018.Yolov3:An incremental improvement[J].arXiv preprint arXiv:1804.02767.
Ren S,He K,Girshick R, et al.2015.Faster r-cnn:Towards real-time object detection with region proposal networks[J].Advances in neural information processing systems,doi:10.48550/arXiv.1506.01497.
Tien B D,Ho T C,Revhaug I, et al.2014.Landslide susceptibility mapping along the national road 32 of Vietnam using GIS-based J48 decision tree classifier and its ensembles[M]//Cartography from pole to pole.Springer,Berlin,Heidelberg,303-317.
Wang C Y,Liao H Y M,Wu Y H, et al.2020.CSPNet:A new backbone that can enhance learning capability of CNN[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops,390-391.
Wang D,He D.2019.Recognition of apple targets before fruits thinning by robot based on R-FCN deep convolution neural network[J].Trans Chin Soc Agric Eng,35:156-163.

备注/Memo

备注/Memo:
收稿日期:2022-04-29.
基金项目:国家重点研发计划课题(2021YFC3000401); 国家自然科学基金项目(41871303); 地质灾害防治与地质环境保护国家重点实验室项目(SKLGP2018Z010); 四川省科技计划项目(2021YFG0365); 四川省自然资源厅(kj-2021-3); 成都市技术创新研发项目(2022-YF05-01090-SN); 成都理工大学研究生质量工程项目(2022YJG022).
第一作者简介:付 饶(2000-),硕士研究生在读,主要从事遥感图像处理、目标识别方面研究.E-mail:2276220889@qq.com.
通讯作者简介:何 敬(1983-),博士,副教授,主要从事无人机影像处理、倾斜三维建模及遥感目标自动识别等方面研究.E-mail:xiao00yao@163.com.
更新日期/Last Update: 2023-03-10