利用Mask R-CNN 辨識建物輪廓與地震風險分析:應用於台灣都會區

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：100

、訪客IP：3.144.42.173

姓名

陳薇安(Wei-An Chen) 查詢紙本館藏

畢業系所

地球科學學系

論文名稱

利用Mask R-CNN 辨識建物輪廓與地震風險分析:應用於台灣都會區
(Detect building footprints by utilizing Mask R-CNN : Application to Metropolitan Taiwan)

相關論文

★ 印尼蘇門答臘地震危害分析	★ 開發深度學習技術的台灣轉換器震動警報模型（TT-SAM）及其應用
★ 應用XGBoost模型建立台灣地殼強地震動衰減式	★ Advancing Probabilistic Seismic Hazard Assessment for Albania (ALhaz2024)
★ 臺灣機率式地震危害評估的地震源模型

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

[檢視]

[下載]

本電子論文使用權限為同意立即開放。
已達開放權限電子全文僅授權使用者為學術研究之目的，進行個人非營利性質之檢索、閱讀、列印。
請遵守中華民國著作權法之相關規定，切勿任意重製、散佈、改作、轉貼、播送，以免觸法。

摘要(中)

本論文開發了一個可自動檢測台灣大都會區的建築物面積的Mask R-CNN 深度學習，並依此建立台灣地震風險模型數據庫的重要元素。為了有效利用 Mask R-CNN 檢測建築物面積，提出了三個步驟：圖像預處理（通過影像銳化處理與從紅外和近紅外波段計算的標準化植被指數來預先處理遙測影像）、模型訓練（以產生通用模型）、後處理（使用遮罩來移除非建築物辨識結果以改善檢測成果）。此模型應用於全台灣，並獲得了約180萬個建物足跡，分別從三個台灣直轄市（新北市、台北市、台中市）計算準確率並得出平均F指數為0.67。另外本研究將建物面積分別與場址效性參數Vs30及斷層最近距離作比較，得出人類活動地區與Vs30與斷層距離均具反比關係，案是人類活動好聚集於場址效應較為顯著之平原區域。本研究的應用成果顯示，深度學習模型提供了更有效的方法來建立建築物資料庫，並且提供了更全面評估自然災害風險的寶貴見解。在實際應用中，本研究所開發的建築物足跡數據庫將為相關部門提供更好的資源分配策略，以應對自然災害風險。

摘要(英)

A deep learning model training is implemented to detect building footprints automatically for the metropolitan area in Taiwan. Building a database is one of the main components to develop a natural hazard risk model for Taiwan. This thesis produced building footprint data through Mask R-CNN, a convolutional neural network (CNN) widely used in image segmentation to predict individual objects. To detect building footprints based on the Mask R-CNN, the thesis is proposed of three procedures: image preprocessing (for image preprocessing, a pan-sharpening multispectral image was obtained from the remote sensing data, and a normalized vegetation index was calculated using the red and near-infrared wavebands to increase data information); model training (for training the deep learning model using the preprocessed data to produce a general model); and post processing (considering masks used to remove non-building-like objects to improve detection results). I applied our model to Taiwan and obtained approximately 1.8 million building footprints with an average F-score of (New Taipei City, Taipei City, Taichung City) of 0.67. In addition, this study compared building area with Vs30 and the nearest distance to the fault, and found that the human activity areas have an inverse relationship with both Vs30 and the fault distance. The deep learning model proposed in this thesis provides a more efficient way to build our building information database, which in turn enhances natural hazard risk assessment. Notably, the application of our model can also be extended to seismic risk assessment, providing a valuable tool for disaster management and prevention efforts.

關鍵字(中)

★ Mask R-CNN
★ 建築物資料庫
★ 地震風險分析

關鍵字(英)

★ Mask R-CNN
★ Building database
★ natural hazard risk assessment

論文目次

CHINESE ABSTRACT i
ENGLISH ABSTRACT ii
ACKNOWLEDGEMENT iii
CONTENTS iv
LIST OF FIGURES vi
LIST OF TABLES ix
Chapter 1 Introduction 1
1.1 Research Background 1
1.2 Previous Work by Others 2
1.3 History of Convolutional Neural Network 3
1.4 Aims and structure of this thesis 5
Chapter 2 Data source: Satellite imagery and training sample 9
2.1 Satellite imagery 9
2.2 Training Data for Building Detection 10
Chapter 3 Deep learning module and detection results 17
3.1 Mask R-CNN 17
3.2 Post Processing 18
3.2.1 Postprocess for clouds 19
3.2.2 Postprocess for riverbank sediments 20
3.2.3 Postprocess for farm lands and seafoam 20
3.2.4 Local model (tailored building detection model designed for Taitung City) 21
3.2.5 County-wise detected result 23
3.3 Accuracy Assessment 24
Chapter 4 Discussion and Convolutional neural network detection results 48
4.1 Discussion 48
4.2 Building database analysis of Vs30 and distance to fault relationship 52
Chapter 5 Conclusions 62
5.1 Summary of the thesis 62
5.2 Future work 62
REFERENCE 64
Appendix 67

參考文獻

[1] E. Fischer, et al., "Objective and Perceived Risk in Seismic Vulnerability Assessment at an Urban Scale," Sustainability, vol. 14, pp. 9380, 2022.

[2] K. Fukushima, "Neocognitron: a self organizing neural network model for a mechanism of pattern recognition unaffected by shift in position," Biological Cybernetics, vol. 36, no. 4, pp. 193-202, 1980.

[3] Y. LeCun, et al., "Backpropagation applied to handwritten zip code recognition," Neural Computation, vol. 1, no. 4, pp. 541–551, 1989.

[4] A. Krizhevsky, I. Sutskever, and G. E. Hinton, "ImageNet classification with deep convolutional neural networks," in Proceedings of the 25th International Conference on Neural Information Processing Systems - Volume 1 (NIPS′12), pp. 1097–1105, Red Hook, NY, USA: Curran Associates Inc., 2012.

[5] K. He, X. Zhang, S. Ren, and J. Sun, "Deep Residual Learning for Image Recognition," in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 770-778, 2015.

[6] C. Ayala, R. Sesma, C. Aranda, and M. Galar, "A Deep Learning Approach to an Enhanced Building Footprint and Road Detection in High-Resolution Satellite Imagery," Remote Sensing, vol. 13, no. 16, pp. 3135, 2021.

[7] S. Ji, Y. Shen, M. Lu, and Y. Zhang, "Building Instance Change Detection from Large-Scale Aerial Images using Convolutional Neural Networks and Simulated Samples," Remote Sensing, vol. 11, no. 11, pp. 1343, 2019.

[8] S. Fujita and X. H. Han, "Cell Detection and Segmentation in Microscopy Images with Improved Mask R-CNN," in Computer Vision – ACCV 2020 Workshops, I. Sato and B. Han, Eds. Lecture Notes in Computer Science, vol. 12628, 2021.

[9] O. Cakiroglu, C. Ozer, and B. Gunsel, "Design of a Deep Face Detector by Mask R-CNN," in 27th Signal Processing and Communications Applications Conference (SIU), Sivas, Turkey, pp. 1-4, 2019.

[10] X. Nie, M. Duan, H. Ding, B. Hu, and E. K. Wong, "Attention Mask R-CNN for ship detection and segmentation from remote sensing images," IEEE Access, vol. 8, pp. 9325-9334, 2020.

[11] D. Tiede, G. Schwendemann, A. Alobaidi, L. Wendt, and S. Lang, "Mask R-CNN-based building extraction from VHR satellite data in operational humanitarian action: An example related to Covid-19 response in Khartoum, Sudan," Transactions in GIS, vol. 25, 2021.

[12] W. Li and Q. Guo, "A New Accuracy Assessment Method for One-Class Remote Sensing Classification," IEEE Transactions on Geoscience and Remote Sensing, vol. 52, pp. 10.110

指導教授

詹忠翰(Chung-Han Chan)

審核日期

2023-6-19

推文