台灣的可耕地面積有限，清查建物的面積有助於了解土地利用的狀況。為了瞭解建物在農業區所佔的總面積，本研究針對農業區進行建物分割。現有的做法之一是透過高解析衛星影像進行人工辨識，此法可以掌握建物的邊界、改善現地調查的不便，然而卻需要大量人力資源的投入。過去的研究顯示，深度學習的方法可以在不同區域的高解析衛星影像自動地分割建物。本研究使用ENVINet5 深度學習模型及Pléiades彩色融合影像進行訓練。因為各地區的建物型態皆不相同，所以研究中所使用的訓練影像包含九個不同的縣市，每張訓練影像的尺寸為2500 × 2500 像素。模型的評估是透過驗證集中的每個像素進行計算，經訓練的模型可以找出84%的建物像素。為了計算建物被預測的數量以及評估建物分割的狀況，本研究將模型所分割出來的建物以多邊形為單位，除了計算其數量，也將其與參考建物以IoU做比較。此評估方法有助於了解建物分割在邊界的實際情況。成果顯示，該模型可以在影像上偵測且分割92%的建物，模型分割出之大部分建物的IoU集中於0.6 到0.9 之間。該模型也以測試集做可轉移性試驗。另外，本研究提出了影像切圖與拼接的方法以處理大範圍的衛星影像。最後，ENVINet5的成果輔助人工辨識建物於75張尺寸為2500 × 2500 像素的影像，並節省了7.3% 的時間成本。

Since arable land area in Taiwan is limited, understanding building area can monitor the situation of land use. The task of building segmentation was carried out in the agricultural land around Taiwan in this study. One of the practical ways is manual labelling from high resolution satellite imagery, which can avoid field investigation and take care of the building borders. However, it is tedious and labor intensive. Past research have shown that deep learning methods are useful to segment buildings from satellite imagery. In this study, ENVINet5 model was trained and utilized to segment buildings from high resolution Pléiades pansharpened imagery. The training images (with the size of 2500 pixels × 2500 pixels) were randomly collected from 9 counties/cities to increase diversity since each county/city has different building patterns. The performance of the trained model was evaluated by every pixel in the validation sets. The pixel-based evaluation shows that the trained model can find 84% of building pixels. To calculate the number of the buildings being segmented and evaluate the segmentation quality of each building segment. The polygon-based evaluation was carried out to count the number of the building segments and compare them with the reference data using IoU. The results showed that the trained model can find 92% of building segments from Pléiades pansharpened agricultural imagery, and most of the building segments have IoU ranging from 0.6 to 0.9. Furthermore, the trained model was validated on the testing images for transferability test. Moreover, an image tiling and stitching technique was proposed to deal with large satellite image for building segmentation. Finally, manual labelling with and without the aid of the deep learning results were compared based on the time consumption. The results showed that with the help of ENVINet5, the time consumption decreased by 7.3% for labelling 75 sub-images (with the size of 2500 pixels × 2500 pixels).

Alshehhi, R., Marpu, P. R., Woon, W. L., & Mura, D. M. (2017). Simultaneous extraction of roads and buildings in remote sensing imagery with convolutional neural networks. ISPRS Journal of Photogrammetry and Remote Sensing. 130, 139-149. https://doi.org/10.1016/j.isprsjprs.2017.05.002
ArcGIS Development Team, 2019. Powerful mapping and analytics software, Version 10.7.1. Esri Corporation. esri.com
Auquilla, A., Heremans, S., Vanegas, P., & Orshoven, J. V. (2014). A Procedure for Semi-automatic Segmentation in OBIA Based on the Maximization of a Comparison Index. Computational Science and Its Applications – ICCSA 2014. 360-375. https://doi.org/10.1007/978-3-319-09144-0_25
Blaschke, T., Lang, S., Lorup, E., Strobl, J., & Zeil, P. (2000). Object-oriented Image Processing in an Integrated GIS/Remote Sensing Environment and Perspectives for Environmental Applications.
Boonpook, W., Tan, Y., Ye, Y., Torteeka, P., Torsri, K., & Dong, S. (2018). A Deep Learning Approach on Building Detection from Unmanned Aerial Vehicle-Based Images in Riverbank Monitoring. Sensors 2018, 18(11), 3921. https://doi.org/10.3390/s18113921
Chen, Y., Tang, L., Yang, X., Bilal, M., & Li, Q. (2020). Object-based multi-modal convolution neural networks for building extraction using panchromatic and multispectral imagery. Neurocomputing. 386, 136-146. https://doi.org/10.1016/j.neucom.2019.12.098
Cheng, D., Liao, R., Fidler, S., & Urtasun, R. (2019). DARNet: Deep Active Ray Network for Building Segmentation. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 7423-7431. https://doi.org/10.1109/CVPR.2019.00761
ENVI Development Team, 2020. Exelis Visual Information Solutions software, Version 5.6. L3HARRIS Geospatial Corporation. www.l3harrisgeospatial.com
Esetlili, T. M., Balcik, F. B., Sanli, F. B., Ustuner, M., Kalkan, K., Goksel, C., Gazioğlu, C., & Kurucu, Y. (2018). Comparison of Object and Pixel-Based Classifications for Mapping Crops Using Rapideye Imagery: A Case Study Of Menemen Plain, Turkey. International Journal of Environment and Geoinformatics. 5(2), 231-243. https://doi.org/10.30897/ijegeo.442002
Gu, J., Wang, Z., Kuen, J., Ma, L., Shahroudy, A., Shuai, B., Liu, T., Wang, X., Wang, G., Cai, J., & Chen, T. (2018). Recent advances in convolutional neural networks. Pattern Recognition. 77, 354-377. https://doi.org/10.1016/j.patcog.2017.10.013
Guo, M., Liu, H., Xu, Y., & Huang, Y. (2020). Building Extraction Based on U-Net with an Attention Block and Multiple Losses. Remote Sensing, 12(9), 1400. https://doi.org/10.3390/rs12091400
Guo, Z., & Du, S. (2017). Mining parameter information for building extraction and change detection with very high-resolution imagery and GIS data. GI Science & Remote Sensing, 54(1), 38-63. https://doi.org/10.1080/15481603.2016.1250328
He, N., Fang, L. & Plaza, A. (2019). Hybrid first and second order attention Unet for building segmentation in remote sensing images. Sci. China Inf. Sci, 63, 140305. https://doi.org/10.1007/s11432-019-2791-7
Hester, D.B., Cakir, H.I., Nelson, S.A.C., & Khorram, S. (2008). Perpixel classification of high spatial resolution satellite imagery for urban land-cover mapping. Photogrammetric Engineering & Remote Sensing. 74(4), 463-471. https://doi.org/10.14358/PERS.74.4.463
Huang, B., Reichman, D., Collins, L., Bradbury, K., & Malof, J.M. (2018). Tiling and Stitching Segmentation Output for Remote Sensing: Basic Challenges and Recommendations. ArXiv. https://arxiv.org/abs/1805.12219
Huang, X., Yuan, W., Li, J., & Zhang, L. (2017). A New Building Extraction Postprocessing Framework for High-Spatial-Resolution Remote-Sensing Imagery. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 10(2), 654-668. https://doi.org/10.1109/JSTARS.2016.2587324
Kastens, J. H., Kastens, T. L., Kastens, D. L. A., Price, K. P., Martinko, E. A., & Lee, R. Y. (2005). Image masking for crop yield forecasting using AVHRR NDVI time series imagery. Remote Sensing of Environment, 341-356. https://doi.org/10.1016/j.rse.2005.09.010
Kaszta, Ż., Kerchove, R. V. D., Ramoelo, A., Cho, M. A., Madonsela, S., Mathieu, R., & Wolff, E. (2016). Seasonal Separation of African Savanna Components Using Worldview-2 Imagery: A Comparison of Pixel- and Object-Based Approaches and Selected Classification Algorithms. Remote Sensing. 8(9), 763. https://doi.org/10.3390/rs8090763
Khan, A., Sohail, A., Zahoora, U. & Qureshi, S. (2020). A survey of the recent architectures of deep convolutional neural networks. Artificial Intelligence Review 53, 5455-5516. https://doi.org/10.1007/s10462-020-09825-6
Kirillov, A., Wu, Y., He, K., & Girshick, R. (2020). PointRend: Image Segmentation
As Rendering. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 9796-9805. https://doi.org/10.1109/CVPR42600.2020.00982
Khosravi, I., Momeni, M., & Rahnemoonfar, M. (2014). Performance Evaluation of Object-based and Pixel-based Building Detection Algorithms from Very High Spatial Resolution Imagery. Photogrammetric Engineering and Remote Sensing. 80(5), 519-528. https://doi.org/10.14358/PERS.80.6.519
Li, W., He, C., Fang, J., Zheng, J., Fu, H., & Yu, L. (2019). Semantic Segmentation-Based Building Footprint Extraction Using Very High-Resolution Satellite Images and Multi-Source GIS Data. Remote Sensing. 11(4), 403. https://doi.org/10.3390/rs11040403
Maltezos, E., Doulamis, N., Doulamis, A., & Ioannidis, C. (2017). Deep convolutional neural networks for building extraction from orthoimages and dense image matching point clouds. Journal of Applied Remote Sensing, 11(4). https://doi.org/10.1117/1.JRS.11.042620
Mou, L., & Zhu, X. X. (2018). RiFCN: Recurrent Network in Fully Convolutional Network for Semantic Segmentation of High Resolution Remote Sensing Images. ArXiv. https://arxiv.org/abs/1805.02091
Nogueira, K., Penatti, O. A., & Santos, J. A. (2017). Towards better exploiting convolutional neural networks for remote sensing scene classification. Pattern Recognition. 61, 539-556. https://doi.org/10.1016/j.patcog.2016.07.001
Prathap, G., & Afanasyev, I. (2018). Deep learning approach for building detection in satellite multispectral imagery. 2018 International Conference on Intelligent Systems (IS), 461-465. https://doi.org/10.1109/IS.2018.8710471
Pu, R., Landry, S., & Yu, Q. (2011). Object-based urban detailed land cover classification with high spatial resolution IKONOS imagery. International Journal of Remote Sensing, 32(12), 3285-3308. https://doi.org/10.1080/01431161003745657
Radoux, J., & Defourny, P. (2008). Quality assessment of segmentation results devoted to object-based classification. Object-Based Image Analysis. 257-271 https://doi.org/10.1007/978-3-540-77058-9_14
Rezatofighi, H., Tsoi, N., Gwak, J., Sadeghian, A., Reid, I., & Savarese S. (2019). Generalized Intersection over Union: A Metric and A Loss for Bounding Box Regression. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 658-666. https://doi.org/10.1109/CVPR.2019.00075.
Ronneberger, O., Fischer, P., & Brox, T. (2015). U-Net: Convolutional Networks for Biomedical Image Segmentation. ArXiv. https://arxiv.org/abs/1505.04597
Shrestha, S., & Vanneschi, L. (2018). Improved Fully Convolutional Network with Conditional Random Fields for Building Extraction. Remote Sensing. 10(7), 1135. https://doi.org/10.3390/rs10071135
Soni, A., Koner, R., & Villuri, V.G.K. (2020). M-UNet: Modified U-Net Segmentation Framework with Satellite Imagery. Proceedings of the Global AI Congress 2019. 47-59. https://doi.org/10.1007/978-981-15-2188-1_4
Sun, W., Chen, B., & Messinger, D. (2013). Nearest-neighbor diffusion-based pan-sharpening algorithm for spectral images. Optical Engineering, 53(1). https://doi.org/10.1117/1.OE.53.1.013107
Trang, N. T. Q., Toan, L. Q., Ai, T. T. H., Giang, N. V., & Hoa, P. V. (2016) Object-Based vs. Pixel-Based Classification of Mangrove Forest Mapping in Vien An Dong Commune, Ngoc Hien District, Ca Mau Province Using VNREDSat-1 Images. Advances in Remote Sensing. 5, 284-295. https://doi.org/10.4236/ars.2016.54022
Ünel, F. Ö., Özkalayci, B. O., & Çiğla, C (2019). The Power of Tiling for Small Object Detection. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 582-591. https://doi.org/10.1109/CVPRW.2019.00084.
Vakalopoulou, M., Karantzalos, K., Komodakis, N., & Paragios, N. (2015). Building detection in very high resolution multispectral data with deep learning features. 2015 IEEE international Geoscience and Remote Sensing Symposium (IGARSS), 1873-1876. https://doi.org/10.1109/IGARSS.2015.7326158
Wang, Z., & Yang, Z. (2020). Review on image-stitching techniques. Multimedia Systems. 26, 413-430. https://doi.org/10.1007/s00530-020-00651-y
Wu, Q., Zhong, R., Zhao, W., Fu, H., & Song, K. (2017). A comparison of pixel-based decision tree and object-based Support Vector Machine methods for land-cover classification based on aerial images and airborne lidar data. International Journal of Remote Sensing. 38(23), 7176-7195. https://doi.org/10.1080/01431161.2017.1371864
Xu, Y., Wu, L., Xie, Z., & Chen, Z. (2018). Building Extraction in Very High Resolution Remote Sensing Imagery Using Deep Learning and Guided Filters. Remote Sensing. 10(1), 144. https://doi.org/10.3390/rs10010144
Yang, H., Wu, P., Yao, X., Wu, Y., Wang, B., & Xu, Y. (2018). Building Extraction in Very High Resolution Imagery by Dense-Attention Networks. Remote Sensing. 10, 1768. https://doi.org/10.3390/rs10111768
Yang, H. L., Yuan, J., Lunga, D., Laverdiere, M., Rose, A., & Bhaduri, B. (2018). Building Extraction at Scale Using Convolutional Neural Network: Mapping of the United States. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. 11(8), 2600-2614. https://doi.org/10.1109/JSTARS.2018.2835377
Yi, Y., Zhang, Z., Zhang, W., Zhang, C., Li, W., & Zhao, T. (2019). Semantic Segmentation of Urban Buildings from VHR Remote Sensing Imagery Using a Deep Convolutional Neural Network. Remote Sensing. 11, 1774. https://doi.org/10.3390/rs11151774
Zhang, L., Wu, J., Fan, Y., Gao, H., & Shao, Y. (2020) An Efficient Building Extraction Method from High Spatial Resolution Remote Sensing Images Based on Improved Mask R-CNN. Sensors. 20(5), 1465. https://doi.org/10.3390/s20051465
Zhao, W., Persello, C., & Stein, A. (2021). Building outline delineation: From aerial images to polygons with an improved end-to-end learning framework. ISPRS Journal of Photogrammetry and Remote Sensing. 175, 119-131. https://doi.org/10.1016/j.isprsjprs.2021.02.014