Intelligent Recognition of Coseismic Landslides Based on MultiU-EGANet Model
-
摘要: 同震滑坡制图在应急救援和灾害评估中具有至关重要的作用.为更好进行滑坡识别，提出了一种新的改进模型——MultiU-EGANet. 该模型以U-Net模型为基线模型，通过引入MultiRes模块，实现对不同尺度特征信息的提取；引入边缘引导注意力模块（edge-guided attention，EGA），通过拉普拉斯算子强化滑坡边界，从而提高模型对边界的分割精度；结合Dice loss和Focal loss构造复合损失函数，进一步增强模型的鲁棒性.基于九寨沟地区滑坡数据进行实验，结果表明改进模型相较于基线模型，滑坡识别精度得到了明显提升. 此外，基于北海道地区滑坡数据进行模型对比实验，结果表明，所提出方法相较于其他现有模型在滑坡识别任务中表现更为优越，F1值分别提升了33.31%、5.45%、2.31%、2.18%.实验结果充分证明了所提出方法在同震滑坡识别中的有效性.Abstract: Coseismic landslide mapping plays a crucial role in emergency response and disaster assessment. To improve landslide identification, this paper proposes a novel and enhanced model, MultiU-EGANet. The model is built upon the U-Net architecture as the baseline, with the introduction of the MultiRes module to extract feature information across multiple scales. Additionally, the Edge-Guided Attention (EGA) module is incorporated to enhance the delineation of landslide boundaries using the Laplace operator, thereby improving the segmentation accuracy at the boundaries. A composite loss function, combining Dice loss and Focal loss, is designed to further enhance the model's robustness. Using landslide data from the Jiuzhaigou area, experimental results demonstrate that the proposed model significantly improves landslide identification accuracy compared to the baseline model. Furthermore, comparative experiments conducted with landslide data from Hokkaido show that the proposed method outperforms existing models in landslide identification tasks, with F1 scores increasing by 33.31%, 5.45%, 2.31%, and 2.18%, respectively. These results validate the effectiveness of the proposed method for coseismic landslide identification.
-
Key words:
- coseismic landslide /
- change detection /
- multiscale /
- edge-guided /
- engineering geology
-
图 3 同震滑坡智能识别流程图
Fig. 3. Flow chart of intelligent identification of coseismic landslides
表 1 本文所使用数据集
Table 1. Datasets used in this study
地区 类型 影像名称 影像时间 分辨率 九寨沟 震前Sentinel-2 Sentinel-2 2017-07-29 10 m 震后Sentinel-2 Sentinel-2 2017-09-07 10 m 震前NDVI Sentinel-2 2017-07-29 10 m 震后NDVI Sentinel-2 2017-09-07 10 m 坡度 SRTM DEM 2000-02-11~2000-02-21 30 m 山体阴影 SRTM DEM 2000-02-11~2000-02-21 30 m 北海道 震前Sentinel-2 Sentinel-2 2017-06-17 10 m 震后Sentinel-2 Sentinel-2 2019-05-23 10 m 震前NDVI Sentinel-2 2017-06-17 10 m 震后NDVI Sentinel-2 2019-05-23 10 m 坡度 SRTM DEM 2000-02-11~2000-02-21 30 m 山体阴影 SRTM DEM 2000-02-11~2000-02-21 30 m 
下载: 导出CSV
表 3 九寨沟区域识别结果
Table 3. Identification results in Jiuzhaigou
模型 Precision Recall F1-score IoU U-Net 0.771 7 0.853 0 0.810 3 0.680 1 MultiU-EGANet 0.805 4 0.863 3 0.833 4 0.713 5
下载: 导出CSV
表 4 北海道地区对比试验结果
Table 4. Results of comparative tests in Hokkaido
模型 Precision Recall F1-score IoU Params FCN 0.544 4 0.755 2 0.632 7 0.462 9 7 916 316 Segnet 0.851 0 0.852 7 0.851 8 0.741 5 7 369 441 Res-UNet 0.867 8 0.876 2 0.872 0 0.772 9 3 269 217 ResU-SENet 0.854 7 0.893 3 0.873 6 0.774 2 3 271 905 MultiU-EGANet 0.878 7 0.894 5 0.886 5 0.796 0 4 718 235
下载: 导出CSV
表 5 九寨沟区域实验结果
Table 5. Experimental results in Jiuzhaigou
模型 Precision Recall F1-score IoU U-Net 0.771 7 0.853 0 0.810 3 0.680 1 MultiU-Net 0.776 7 0.862 8 0.817 5 0.690 7 U-EGANet 0.783 7 0.854 8 0.817 7 0.687 5 MultiU-EGANet 0.805 4 0.863 3 0.833 4 0.713 5
下载: 导出CSV
表 6 九寨沟损失函数权重选择
Table 6. Weight selection of loss function in Jiuzhaigou
a b Precision Recall F1-score IoU 1 0 0.787 1 0.874 5 0.828 5 0.704 8 0.9 0.1 0.790 5 0.865 9 0.826 5 0.701 6 0.8 0.2 0.7837 0.859 0 0.819 6 0.690 3 0.7 0.3 0.805 4 0.863 3 0.833 4 0.713 5 0.6 0.4 0.792 8 0.851 9 0.821 3 0.694 5 0.5 0.5 0.779 0 0.890 3 0.830 9 0.710 3 0.4 0.6 0.767 4 0.892 6 0.825 3 0.701 9 0.3 0.7 0.783 4 0.851 7 0.816 1 0.685 3 0.2 0.8 0.784 6 0.864 2 0.822 5 0.697 5 0.1 0.9 0.760 5 0.861 6 0.807 9 0.673 7 0 1 0.803 2 0.820 8 0.811 9 0.680 0
下载: 导出CSV
表 7 北海道损失函数权重选择
Table 7. Weight selection of loss function in Hokkaido
a b Precision Recall F1-score IoU 1 0 0.863 1 0.908 3 0.885 1 0.793 7 0.9 0.1 0.878 7 0.894 5 0.886 5 0.796 0 0.8 0.2 0.857 6 0.905 4 0.880 9 0.786 6 0.7 0.3 0.866 8 0.902 2 0.884 2 0.792 0 0.6 0.4 0.859 0 0.903 5 0.880 7 0.786 5 0.5 0.5 0.860 1 0.909 2 0.884 0 0.791 8 0.4 0.6 0.855 2 0.908 8 0.881 2 0.787 2 0.3 0.7 0.859 3 0.915 0 0.886 3 0.795 6 0.2 0.8 0.864 3 0.900 4 0.882 0 0.788 8 0.1 0.9 0.872 9 0.895 1 0.883 9 0.791 6 0 1 0.878 5 0.881 5 0.880 0 0.785 7
下载: 导出CSV
表 8 不同源域的模型在帕卢地区的滑坡识别结果
Table 8. Landslide identification results of models with different source domains in Palu
源域 泛化域 模型 Precision Recall F1-score IoU 九寨沟 帕卢 U-Net 0.106 1 0.020 5 0.032 0 0.016 7 九寨沟 MultiU-EGANet 0.159 4 0.160 8 0.152 9 0.083 9 北海道 U-Net 0.966 0 0.246 4 0.376 1 0.241 7 北海道 MultiU-EGANet 0.931 2 0.543 2 0.681 3 0.522 2
下载: 导出CSV
-
Badrinarayanan, V., Kendall, A., Cipolla, R., 2017. SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(12): 2481-2495 Borgwardt, K. M., Gretton, A., Rasch, M. J., et al., 2006. Integrating Structured Biological Data by Kernel Maximum Mean Discrepancy. Bioinformatics, 22(14): e49-e57. https://doi.org/10.1093/bioinformatics/btl242 Bragagnolo, L., Rezende, L. R., da Silva, R. V., et al., 2021. Convolutional Neural Networks Applied to Semantic Segmentation of Landslide Scars. CATENA, 201: 105189. https://doi.org/10.1016/j.catena.2021.105189 Bui, N. T., Hoang, D. H., Nguyen, Q. T., et al., 2024. MEGANet: Multi-Scale Edge-Guided Attention Network for Weak Boundary Polyp Segmentation. 2024 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV). January 3-8, 2024, Waikoloa, HI, USA. IEEE: 7970-7979. https://doi.org/10.1109/WACV57701.2024.00780 Chang, M., Zhou, Y., Zhou, C., et al., 2021. Coseismic Landslides Induced by the 2018 Mw 6.6 Iburi, Japan, Earthquake: Spatial Distribution, Key Factors Weight, and Susceptibility Regionalization. Landslides, 18(2): 755-772. https://doi.org/10.1007/s10346-020-01522-3 Chen, B., Song, C., Chen, Y., et al., 2025. Emergency Identification and Influencing Factor Analysis of Coseismic Landslides and Building Damages Induced by the 2023 Ms 6.2 Jishishan (Gansu, China) Earthquake. Geomatics and Information Science of Wuhan University, 50(2): 322-332(in Chinese with English abstract). Chen, L. Q., Zhao, C. Y., Ren, C. F., et al., 2020. Monitoring the Jianshanying Landslide in a Karst Mountainous Area of Guizhou by Optical Remote Sensing. Carsologica Sinica, 39(4): 518-523(in Chinese with English abstract). Dai, L. X., Fan, X. M., Wang, X., et al., 2023. Coseismic Landslides Triggered by the 2022 Luding Ms6.8 Earthquake, China. Landslides, 20(6): 1277-1292. https://doi.org/10.1007/s10346-023-02061-3 Dai, L. X., Xu, Q., Fan, X. M., et al., 2017. A Preliminary Study on Spatial Distribution Patterns of Landslides Triggered by Jiuzhaigou Earthquake in Sichuan on August 8TH, 2017 and Their Susceptibility Assessment. Journal of Engineering Geology, 25(4): 1151-1164(in Chinese with English abstract). Dou, J., Xiang, Z. L., Xu, Q., et al., 2023. Application and Development Trend of Machine Learning in Landslide Intelligent Disaster Prevention and Mitigation. Earth Science, 48(5): 1657-1674(in Chinese with English abstract). Emberson, R., Kirschbaum, D., Stanley, T., 2021. Global Connections between El Nino and Landslide Impacts. Nature Communications, 12: 2262. https://doi.org/10.1038/s41467-021-22398-4 Fan, X. M., Scaringi, G., Xu, Q., et al., 2018. Coseismic Landslides Triggered by the 8th August 2017 Ms 7.0 Jiuzhaigou Earthquake (Sichuan, China): factors Controlling Their Spatial Distribution and Implications for the Seismogenic Blind Fault Identification. Landslides, 15(5): 967-983. https://doi.org/10.1007/s10346-018-0960-x Ghorbanzadeh, O., Meena, S. R., Shahabi Sorman Abadi, H., et al., 2021. Landslide Mapping Using Two Main Deep-Learning Convolution Neural Network Streams Combined by the Dempster-Shafer Model. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 14: 452-463 Guo, Z. L., Wen, Y. M., Xu, G. Y., et al., 2019. Fault Slip Model of the 2018 Mw 6.6 Hokkaido Eastern Iburi, Japan, Earthquake Estimated from Satellite Radar and GPS Measurements. Remote Sensing, 11(14): 1667. https://doi.org/10.3390/rs11141667 Hacıefendioğlu, K., Demir, G., Başağa, H. B., 2021. Landslide Detection Using Visualization Techniques for Deep Convolutional Neural Network Models. Natural Hazards, 109(1): 329-350. https://doi.org/10.1007/s11069-021-04838-y He, K. M., Zhang, X. Y., Ren, S. Q., et al., 2016. Identity Mappings in Deep Residual Networks. Computer Vision - ECCV 2016. Cham: Springer International Publishing: 630-645. https://doi.org/10.1007/978-3-319-46493-0_38 Ibtehaz, N., Rahman, M. S., 2020. MultiResUNet: Rethinking the U-Net Architecture for Multimodal Biomedical Image Segmentation. Neural Networks, 121: 74-87. https://doi.org/10.1016/j.neunet.2019.08.025 Ji, S. P., Yu, D. W., Shen, C. Y., et al., 2020. Landslide Detection from an Open Satellite Imagery and Digital Elevation Model Dataset Using Attention Boosted Convolutional Neural Networks. Landslides, 17(6): 1337-1352. https://doi.org/10.1007/s10346-020-01353-2 Jiang, H. W., Peng, M., Zhong, Y. J., et al., 2022. A Survey on Deep Learning-Based Change Detection from High-Resolution Remote Sensing Images. Remote Sensing, 14(7): 1552. https://doi.org/10.3390/rs14071552 Keefer, D. K., 1984. Landslides Caused by Earthquakes. Geological Society of America Bulletin, 95(4): 406. https://doi.org/10.1130/0016-7606(1984)95<406:LCBE>2.0.CO;2 doi: 10.1130/0016-7606(1984)95<406:LCBE>2.0.CO;2 Lan, H. X., Li, L. P., Zhang, Y. S., et al., 2013. Risk Assessment of Debris Flow in Yushu Seismic Area in China: a Perspective for the Reconstruction. Natural Hazards and Earth System Sciences, 13(11): 2957-2968. https://doi.org/10.5194/nhess-13-2957-2013 Li, Y. L., Ding, M. T., Zhang, Q., et al., 2024. Old Landslide Detection Using Optical Remote Sensing Images Based on Improved YOLOv8. Applied Sciences, 14(3): 1100. https://doi.org/10.3390/app14031100 Lin, T. Y., Goyal, P., Girshick, R., et al., 2020. Focal Loss for Dense Object Detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 42(2): 318-327. https://doi.org/10.1109/TPAMI.2018.2858826 Liu, J., Wu, Y. M., Gao, X., et al., 2022. Image Recognition of Coseismic Landslide Based on GEE and U-Net Neural Network. Journal of Geo-Information Science, 24(7): 1275-1285(in Chinese with English abstract). Liu, P., Wei, Y. M., Wang, Q. J., et al., 2020. Research on Post-Earthquake Landslide Extraction Algorithm Based on Improved U-Net Model. Remote Sensing, 12(5): 894. https://doi.org/10.3390/rs12050894 Lu, P., Shi, W. Y., Wang, Q. M., et al., 2021. Co-Seismic Landslide Mapping Using Sentinel-2 10m Fused NIR Narrow, Red-Edge, and SWIR Bands. Landslides, 18(6): 2017-2037. https://doi.org/10.1007/s10346-021-01636-2 Milletari, F., Navab, N., Ahmadi, S. A., 2016. V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation. 2016 Fourth International Conference on 3D Vision (3DV). October 25-28, 2016, Stanford, CA, USA. IEEE: 565-571. https://doi.org/10.1109/3DV.2016.79 Ronneberger, O., Fischer, P., Brox, T., 2015. U-Net: Convolutional Networks for Biomedical Image Segmentation, in: Navab, N., Hornegger, J., Wells, W. M., eds., Medical Image Computing and Computer-Assisted Intervention: MICCAI 2015, Springer International Publishing, Cham, 234-241. Shelhamer, E., Long, J., Darrell, T., 2017. Fully Convolutional Networks for Semantic Segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(4): 640-651. https://doi.org/10.1109/TPAMI.2016.2572683 Tang, X. G., Wang, L. J., Wang, H. Y., et al., 2024. Predicted Climate Change Will Increase Landslide Risk in Hanjiang River Basin, China. Journal of Earth Science, 35(4): 1334-1354. https://doi.org/10.1007/s12583-021-1511-2 Wang, X., Fan, X. M., Xu, Q., et al., 2022. Change Detection-Based Co-Seismic Landslide Mapping through Extended Morphological Profiles and Ensemble Strategy. ISPRS Journal of Photogrammetry and Remote Sensing, 187: 225-239. https://doi.org/10.1016/j.isprsjprs.2022.03.011 Woo, S., Park, J., Lee, J. Y., et al., 2018. CBAM: Convolutional Block Attention Module. In: Ferrari, V., Hebert, M., Sminchisescu, C., eds., Computer Vision: ECCV 2018, PT VII, Lecture Notes in Computer Science, Presented at the 15th European Conference on Computer Vision (ECCV), Springer International Publishing Ag, Cham, 3-19. Xu, Q., Dong, X. J., Li, W. L., 2019. Integrated Space-Air-Ground Early Detection, Monitoring and Warning System for Potential Catastrophic Geohazards. Geomatics and Information Science of Wuhan University, 44(7): 957-966(in Chinese with Englis abstract). Yamagishi, H., Yamazaki, F., 2018. Landslides by the 2018 Hokkaido Iburi-Tobu Earthquake on September 6. Landslides, 15(12): 2521-2524. https://doi.org/10.1007/s10346-018-1092-z Yang, J., Ding, M. T., Huang, W. B., et al., 2024. A Generalized Deep Learning-Based Method for Rapid Co-Seismic Landslide Mapping. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 17: 16970-16983 Zhang, S., Li, R., Wang, F. W., et al., 2019. Characteristics of Landslides Triggered by the 2018 Hokkaido Eastern Iburi Earthquake, Northern Japan. Landslides, 16(9): 1691-1708. https://doi.org/10.1007/s10346-019-01207-6 Zhang, Z. X., Liu, Q. J., Wang, Y. H., 2018. Road Extraction by Deep Residual U-Net. IEEE Geoscience and Remote Sensing Letters, 15(5): 749-753. https://doi.org/10.1109/LGRS.2018.2802944 Zhao, W., Li, A. N., Nan, X., et al., 2017. Postearthquake Landslides Mapping from Landsat-8 Data for the 2015 Nepal Earthquake Using a Pixel-Based Change Detection Method. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 10(5): 1758-1768. https://doi.org/10.1109/JSTARS.2017.2661802 陈博, 宋闯, 陈毅, 等, 2025. 2023年甘肃积石山Ms 6.2地震同震滑坡和建筑物损毁情况应急识别与影响因素研究. 武汉大学学报(信息科学版), 50(2): 322-332. 陈立权, 赵超英, 任超锋, 等, 2020. 光学遥感用于贵州发耳镇尖山营滑坡监测研究. 中国岩溶, 39(4): 518-523. 戴岚欣, 许强, 范宣梅, 等, 2017. 2017年8月8日四川九寨沟地震诱发地质灾害空间分布规律及易发性评价初步研究. 工程地质学报, 25(4): 1151-1164. 窦杰, 向子林, 许强, 等, 2023. 机器学习在滑坡智能防灾减灾中的应用与发展趋势. 地球科学, 48(5): 1657. doi: 10.3799/dqkx.2022.419 刘佳, 伍宇明, 高星, 等, 2022. 基于GEE和U-Net模型的同震滑坡识别方法. 地球信息科学学报, 24(7): 1275-1285. 许强, 董秀军, 李为乐, 2019. 基于天-空-地一体化的重大地质灾害隐患早期识别与监测预警. 武汉大学学报(信息科学版), 44(7): 957-966. -
点击查看大图
图(13) / 表(8)
计量
- 文章访问数: 724
- HTML全文浏览量: 440
- PDF下载量: 38
- 被引次数: 0
