使用Sentinel -2 影像提出空間、光譜與時間的深度學習架構製作佛羅里達州西南部於2017年受艾瑪颶風影響之紅樹林退化圖

簡易檢索 / 詳目顯示

回結果列表

研究生：	杰努庭 Ilham Jamaluddin
論文名稱：	使用Sentinel -2 影像提出空間、光譜與時間的深度學習架構製作佛羅里達州西南部於2017年受艾瑪颶風影響之紅樹林退化圖 Proposed Spatial-Spectral-Temporal Deep Learning Architecture for Mangrove Degradation Mapping Affected by Hurricane Irma 2017 Using Sentinel-2 Data in Southwest Florida
指導教授：	陳映濃 Chen, Ying-Nong
口試委員:
學位類別：	碩士 Master
系所名稱：	太空及遙測研究中心 - 遙測科技碩士學位學程 Master of Science Program in Remote Sensing Science and Technology
論文出版年：	2022
畢業學年度：	110
語文別：	英文
論文頁數：	59
中文關鍵詞：	Sentinel-2 、紅樹林、退化、颶風、卷積、LSTM 、深度學習
外文關鍵詞：	Sentinel-2, mangroves, degradation, hurricane, convolutional, LSTM, deep learning
相關次數：	點閱：24 下載：0
分享至:	分享至facebook 分享至twitter

查詢本校圖書館目錄查詢臺灣博碩士論文知識加值系統勘誤回報

紅樹林是生長在熱帶和亞熱帶氣候區周圍潮間帶的獨特植被，對人類及周邊生態系統有諸多益處（如海岸防護植被、儲碳植被等），紅樹林地圖是了解紅樹林狀況和狀況的重要訊息，衛星影像是廣泛用於紅樹林測繪的數據之一。自然災害事件可能導致紅樹林退化包括颶風事件。在2017 年颶風艾瑪襲擊了佛羅里達州西南部沿海地區導致紅樹林退化。本研究的目的是受颶風艾瑪影響的紅樹林退化地圖提出空間-光譜-時間的深度學習 (DL) 架構。在這項研究中，原本研究區域內的紅樹林區域在颶風事件之後已經退化，透過艾瑪颶風前後的衛星圖像之間的關係可以更深入地了解退化的紅樹林地區。本研究使用免費提供的 Sentinel-2 圖像。所提出的 DL 架構由兩個子模型組成：第一個子模型是 pre-post 深度特徵提取器，利用卷積長短期記憶提取颶風事件前後衛星圖像的時空關係（ConvLSTM）和第二個子模型是典型的全卷積網絡（FCN）分類，將從第一個子模型中提取的特徵分為三類。根據實驗結果，該模型的算法輸出的完整和退化紅樹林類的交叉聯合（IoU）得分分別為96.47％和96.82％，而地圖用戶對生成的地圖的完整和退化紅樹林類的準確度分別為分別為 96.80% 和 95.40%。所提出的模型比其他現有的 FCN 模型（U-Net、LinkNet、FPN 和 FC-DenseNet）取得了更好的結果

Mangroves are unique vegetation that grows in the intertidal zone around tropical and subtropical climate areas and has many benefits for humans and the surrounding ecosystem (e.g. coastal protection vegetation, carbon storage vegetation, etc.). Mangrove map is important information to know the condition and status of mangrove forests. Satellite imagery is one of the data that is widely used for mangrove mapping. Some natural hazard events can cause mangrove degradation including hurricane events. In 2017, Hurricane Irma hit the southwest Florida coastal zone and caused mangrove degradation. The goal of this study is to propose spatial-spectral-temporal deep learning (DL) architecture for mangrove degradation mapping that was affected by Hurricane Irma. In this study, the degraded mangroves are the mangrove area before the hurricane but were degraded after the hurricane event. The relationship between satellite imagery before and after Hurricane Irma can provide a deeper understanding of degraded mangrove areas. This study used freely available Sentinel-2 imagery. The proposed DL architecture consists of two sub-models: the first sub-model is the pre-post deep feature extractor to extract the spatial-spectral-temporal relationship of satellite imagery before and after the hurricane event by using convolutional long-short term memory (ConvLSTM) and the second sub-model is typical fully convolutional network (FCN) classification to classify the extracted features from the first sub-model into three classes. Based on experiment results, the algorithm output intersection over union (IoU) score of intact and degraded mangrove classes from the proposed model are 96.47% and 96.82%, respectively, while the map user’s accuracy of intact and degraded mangrove class from the produced map are 96.80% and 95.40%, respectively. The proposed model achieved better results than other existing FCN models (U-Net, LinkNet, FPN, and FC-DenseNet).

摘要    i
Abstract    ii
Acknowledgment    iii
Table of Contents    iv
List of Figures    vi
List of Tables    vii
CHAPTER I INTRODUCTION    1
1. Background    1
2. Challenge and Objectives    3
CHAPTER II LITERATURE REVIEW    4
1 Mangroves    4
2. Mangrove Degradation    5
3. Remote Sensing for Mangrove    5
4. Deep Learning for Remote Sensing    7
5. Related Works    8
CHAPTER III Study Area and Methods    9
1. Study Area    9
2. Dataset    10
2.1. Sentinel-2 Pre-Processing    10
2.2. Input Data for Model    13
2.3. Map Accuracy Assessment Reference Samples    16
3. Methods    17
4. Proposed Deep Learning Architecture    18
4.1. Pre-Post Deep Feature Extractor    19
4.2. FCN Classifier    21
5. Evaluation Assessments    22
5.1. Algorithm Output Evaluation    23
5.2.  Map Accuracy Assessment    23
6. Implementation Details    24
CHAPTER IV Results and Discussion    26
1. Proposed Deep Learning Architecture Result    26
2. Ablation Experiments    29
2.1. Effects of Spectral Indices    29
2.2. Effects of Each Extractor Part    31
3. Comparison with Existing Architecture    32
4. Analysis of Intact and Degraded Mangrove Map    37
5. Discussion    38
CHAPTER V Conclusions and Future Work    43
1. Conclusion    43
2. Future Work    43
References    44

                                

[1] Giri, C.; Ochieng, E.; Tieszen, L.L.; Zhu, Z.; Singh, A.; Loveland, T.; Masek, J.; Duke, N. Status and Distribution of Mangrove Forests of The World Using Earth Observation Satellite Data. Glob. Ecol. Biogeogr. 2011, 20, 154–159.
[2] Tang, W.; Zheng, M.; Zhao, X.; Shi, J.; Yang, J.; Trettin, C. Big Geospatial Data Analytics for Global Mangrove Biomass and Carbon Estimation. Sustainability 2018, 10, 472
[3] Barua, P.; Rahman, S. Sustainable Livelihood of Vulnerable Communities in Southern Coast of Bangladesh through the Uti-lization of Mangroves. Asian J. Water Environ. Pollut. 2019, 16, 59–67
[4] Shahbudin, S.; Zuhairi, A.; Kamaruzzaman, B.Y. Impact of Coastal Development on Mangrove Cover in Kilim river, Langkawi Island, Malaysia. J. For. Res. 2012, 23, 185–190
[5] Carugati, L.; Gatto, B.; Rastelli, E.; Martire, M.L.; Coral, C.; Greco, S.; Danovaro, R. Impact of Mangrove Forests Degradation on Biodiversity and Ecosystem Functioning. Sci. Rep. 2018, 8, 13298.
[6] Hamilton, S.E.; Casey, D. Creation of a High Spatio-Temporal Resolution Global Database of Continuous Mangrove Forest Cover for the 21st Century (CGMFC-21). Glob. Ecol. Biogeogr. 2016, 25, 729–738
[7] Zhen, J.; Liao, J.; Shen, G. Mapping Mangrove Forests of Dongzhaigang Nature Reserve in China Using Landsat 8 and Radarsat-2 Polarimetric SAR Data. Sensors 2018, 18, 4012
[8] Spencer, T.; Schuerch, M.; Nicholls, R.J.; Hinkel, J.; Lincke, D.; Vafeidis, A.T.; Reef, R.; McFadden, L.; Brown, S. Global coastal wetland change under sea-level rise and related stresses: The DIVA Wetland Change Model. Glob. Planet. Chang. 2016, 139, 15–30
[9] Gandhi, S.; Jones, T. Identifying Mangrove Deforestation Hotspots in South Asia, Southeast Asia and Asia-Pacific. Remote Sens. 2019, 11, 728
[10] Rivera-Monroy, V.H.; Danielson, T.M.; Castañeda-Moya, E.; Marx, B.D.; Travieso, R.; Zhao, X.; Farfan, L.M. Long-term demography and stem productivity of Everglades mangrove forests (Florida, USA): Resistance to hurricane disturbance. For. Ecol. Manag. 2019, 440, 79–91.
[11] McCarthy, M.J.; Jessen, B.; Barry, M.J.; Figueroa, M.; McIntosh, J.; Murray, T.; Schmid, J.; Muller-Karger, F.E. Automated High-Resolution Time Series Mapping of Mangrove Forests Damaged by Hurricane Irma in Southwest Florida. Remote Sens. 2020, 12, 1740.
[12] Diniz, C.; Cortinhas, L.; Nerino, G.; Rodrigues, J.; Sadeck, L.; Adami, M.; Souza-Filho, P.W.M. Brazilian Mangrove Status: Three Decades of Satellite Data Analysis. Remote Sens. 2019, 11, 808.
[13] Mondal, P.; Liu, X.; Fatoyinbo, T.E.; Lagomasino, D. Evaluating Combinations of Sentinel-2 Data and Machine-Learning Algorithms for Mangrove Mapping in West Africa. Remote Sens. 2019, 11, 2928.
[14] Kamal, M.; Phinn, S.; Johansen, K. Object-Based Approach for Multi-Scale Mangrove Composition Mapping Using Mul-ti-Resolution Image Datasets. Remote Sens. 2015, 7, 4753–4783.
[15] Wan, L.; Zhang, H.; Lin, G.; Lin, H. A small-patched convolutional neural network for mangrove mapping at species level using high-resolution remote-sensing image. Ann. GIS 2019, 25, 45–55.
[16] Guo, M.; Yu, Z.; Xu, Y.; Huang, Y.; Li, C. ME-Net: A Deep Convolutional Neural Network for Extracting Mangrove Using Sentinel-2A Data. Remote Sens. 2021, 13, 1292.
[17] Guo, Y.; Liao, J.; Shen, G. Mapping Large-Scale Mangroves along the Maritime Silk Road from 1990 to 2015 Using a Novel Deep Learning Model and Landsat Data. Remote Sens. 2021, 13, 245.
[18] McCarthy, M.J.; Jessen, B.; Barry, M.J.; Figueroa, M.; McIntosh, J.; Murray, T.; Schmid, J.; Muller-Karger, F.E. Mapping hurricane damage: A comparative analysis of satellite monitoring methods. Int. J. Appl. Earth Obs. Geoinf. 2020, 91, 102134.
[19] Lee, C.K.F.; Duncan, C.; Nicholson, E.; Fatoyinbo, T.E.; Lagomasino, D.; Thomas, N.; Worthington, T.A.; Murray, N.J. Mapping the Extent of Mangrove Ecosystem Degradation by Integrating an Ecological Conceptual Model with Satellite Data. Remote Sens. 2021, 13, 2047.
[20] Shi, X.; Chen, Z.; Wang, H.; Yeung, D-Y. Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting. Comput. Vis. Pattern Recognit. 2015, 2, 802–810.
[21] Lugo, A.E; and Snedaker, S.C. The Ecology of Mangroves. Ann. Rev. Ecol. Syst. 1974, 5: 39‐64.
[22] Biswas, S.R; Mallik, A.U; Choudhury, J.K; Nishat, A. A unified framework for the restoration of Southeast Asian mangroves—bridging ecology, society and economics. Wetlands Ecol Manage. 2009, 17, 365–383.
[23] Sippo, J.Z.; Lovelock, C.E.; Santos, I.R.; Sanders, C.J.; Maher, D.T. Mangrove mortality in a changing climate: An overview. Estuar. Coast. Shelf Sci. 2018, 215, 241–249.
[24] Lillesand, T.M; Kiefer, R.W; Chipman, J.W; Remote Sensing and Image Interpretation. 2008. New York: John Willey & Sons Inc.
[25] Huete, A.R and Glenn, E.P. Remote Sensing of Ecosystem Structure and Function. 2011. Boca Raton: CRC Press, Advance in Environment Remote Sensing.
[26] Hoffer, R. M; Davis, S. M; Landgrebe, D. A; Phillips, T. L; Swain, P. H; Lindenlaub, J. C; Silva, L. F. Remote Sensing: The Quantitative Approach. 1928. New York, McGraw-Hill International Book Co., 1978. 405 p.
[27] Roman, A and Ursu, T. Multispektral Satellite Imagery and Airborne Laser Scanning Techniques for the Detection of Archaeological Vegetation Marks. 2016. ISBN 978-606-543-787-6.
[28] LeCun, Y.; Bengio, Y.; Hinton, G. Deep learning. Nature. 2015, 521, 436.
[29] Neupane, B.; Horanont, T.; Aryal, J. Deep Learning-Based Semantic Segmentation of Urban Features in Satellite Images: A Review and Meta-Analysis. Remote Sens. 2021, 13, 808.
[30] Hoeser, T.; Bachofer, F.; Kuenzer, C. Object Detection and Image Segmentation with Deep Learning on Earth Observation Data: A Review—Part II: Applications. Remote Sens. 2020, 12, 3053.
[31] Chen, Y.-N.; Thaipisutikul, T.; Han, C.-C.; Liu, T.-J.; Fan, K.-C. Feature Line Embedding Based on Support Vector Machine for Hyperspectral Image Classification. Remote Sens. 2021, 13, 130.
[32] Long, J.; Shelhamer, E.; Darrell, T. Fully Convolutional Networks for Semantic Segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 2014, 39, 640–651.
[33] Chaurasia, A.; Culurciello, E. LinkNet: Exploiting Encoder Representations for Efficient Semantic Segmentation. In Pro-ceedings of the 2017 IEEE Visual Communications and Image Processing (VCIP), St. Petersburg, FL, USA, 10–13 December 2017.
[34] Ronneberger, O.; Fischer, P.; Brox, T. U-Net: Convolutional Networks for Biomedical Image Segmentation. In Proceedings of the Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015, Munich, Germany, 5–9 October 2015; Springer International Publishing: Cham, Switzerland, 2015; pp. 234–241.
[35] Jégou, S.; Drozdzal, M.; Vazquez, D.; Romero, A.; Bengio, Y. The One Hundred Layers Tiramisu: Fully Convolutional DenseNets for Semantic Segmentation. In Proceedings of the IEEE conference on computer vision and pattern recognition workshops, Honolulu, HI, USA 21–26 July 2017.
[36] Simonyan, K.; Zisserman, A. Very Deep Convolutional Networks for Large-Scale Image Recognition. arXiv 2014, arXiv:1409.1556.
[37] Krizhevsky, A.; Sutskever, I.; Hinton, G.E. ImageNet Classification with Deep Convolutional Neural Networks. In Pro-ceedings of the Advances in Neural Information Processing Systems 25, Red Hook, NY, USA, 3–6 December 2012; pp. 1097–1105.
[38] Huang, G.; Liu, Z.; Van Der Maaten, L.; Weinberger, K.Q. Densely Connected Convolutional Networks. In Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, 21–26 July 2017; pp. 2261–2269.
[39] He, K.; Zhang, X.; Ren, S.; Sun, J. Deep Residual Learning for Image Recognition. In Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 26 June–1 July 2016; pp. 770–778.
[40] Gallego, A.-J.; Gil, P.; Pertusa, A.; Fisher, R.B. Semantic Segmentation of SLAR Imagery with Convolutional LSTM Sectional AutoEncoders. Remote Sens. 2019, 11, 1402.
[41] Pfeuffer, A.; Schulz, K.; Dietmayer, K. Semantic Segmentation of Video Sequences with Convolutional LSTMs. Comput. Vis. Pattern Recognit. 2019, 1, 1441–1447.
[42] Nabavi, S.S.; Rochan, M.; Wang, Y. Future Semantic Segmentation with Convolutional LSTM. arXiv 2018, arXiv:1807.07946.
[43] Worthington, T.A.; zu Ermgassen, P.S.E.; Friess, D.A.; Krauss, K.W.; Lovelock, C.E.; Thorley, J.; Tingey, R.; Woodroffe, C.D.; Bunting, P.; Cormier, N.; et al. A global biophysical typology of mangroves and its relevance for ecosystem structure and deforestation. Sci. Rep. 2020, 10, 14652.
[44] Jamaluddin, I.; Thaipisutikul, T.; Chen, Y.-N.; Chuang, C.-H.; Hu, C.-L. MDPrePost-Net: A Spatial-Spectral-Temporal Fully Convolutional Network for Mapping of Mangrove Degradation Affected by Hurricane Irma 2017 Using Sentinel-2 Data. Remote Sens. 2021, 13, 5042.
[45] Tucker, C.J. Red and photographic infrared linear combinations for monitoring vegetation. Remote Sens. Environ. 1979, 8, 127–150.
[46] McFeeters, S.K. The use of the Normalized Difference Water Index (NDWI) in the delineation of open water features. Int. J. Remote Sens. 1996, 17, 1425–1432.
[47] Gupta, K.; Mukhopadhyay, A.; Giri, S.; Chanda, A.; Datta Majumdar, S.; Samanta, S.; Mitra, D.; Samal, R.N.; Pattnaik, A.K.; Hazra, S. An index for discrimination of mangroves from non-mangroves using LANDSAT 8 OLI imagery. MethodsX. 2018, 5, 1129–1139.
[48] Shi, T.; Liu, J.; Hu, Z.; Liu, H.; Wang, J.; Wu, G. New spectral metrics for mangrove forest identification. Remote. Sens. Lett. 2016, 7, 885–894.
[49] Xu, H. Modification of normalized difference water index (NDWI) to enhance open water features in remotely sensed imagery. Int. J. Remote Sens. 2006, 27, 3025–3033.
[50] Szegedy, C.; Liu, Wei.; Jia, Y.; Sermanet, P.; Reed, S.; Anguelov, D.; Erhan, D.; Vanhoucke, V.; Rabinovich, A. Going Deeper with Convolutions. In Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Bos-ton, MA, USA, 7–12 June 2015.
[51] Rezatofighi, S.H.; Tsoi, N.; Gwak, J.; Sadeghian, A.; Reid, I.D.; Savarese, S. Generalized Intersection over Union: A Metric and A Loss for Bounding Box Regression. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, CA, USA, 15–20 June 2019.
[52] Saferbekov, S.; Iglovikov, V.; Buslaev, A.; Shvets, A. Feature Pyramid Network for Multi-Class Land Segmentation. Computer Vision and Pattern Recognition. 2018, 2.
[53] Chicco, D.; Warrens, M.J.; Jurman, G. The Matthews Correlation Coefficient (MCC) is More Informative Than Cohen’s Kappa and Brier Score in Binary Classification Assessment. IEEE Access. 2021, 9, 78368–78381.

簡易檢索 / 詳目顯示

相關論文