Automatic Building Extraction and Regularization from Image Matching Based Point Cloud

Yıl 2022, Cilt: 4 Sayı: 1, 25 - 35, 15.06.2022

Buray Karslı , Ferruh Yılmaztürk

https://doi.org/10.51489/tuzal.1098240

Öz

Manually determining building layer boundaries with classical or remote sensing data is a time-consuming and effort-intensive process. Point clouds produced by matching from images contain dense and high-accuracy 3D information. Automatic extraction of buildings from 3D point clouds is a difficult problem in terms of geometric irregularities and the density and precision of the point cloud. In this study, the improved Octree (I-Octree) approach was developed by automating the voxel-based octree method, and automatic extraction and regularization of building details on point clouds produced from images are aimed. Point clouds were produced in study area (Elazig region), ground and above ground objects were sorted by SMRF, building objects are classified by removing noise with DBSCAN algorithm, and Octree and I-Octree methods were applied to the classified objects, then the edges of the building details are smoothed with the ABORE method. Automatically extracted building data were validated with pixel-based completeness, accuracy, quality, and F-score metrics with the support of reference map containing the study area. Validation results were obtained for each metric above 94%. It was concluded that the I-Octree approach developed can contribute to a fast and inexpensive map production process at the point of extracting the building details.

Anahtar Kelimeler

Photogrammetry, point cloud, building extraction, I-Octree, digitalization

Görüntü Eşleştirme Kaynaklı Nokta Bulutu Üzerinden Otomatik Bina Çıkarımı ve Düzgünleştirme

Öz

Harita bileşeni bina katmanı sınırlarının klasik veya uzaktan algılama verilerinden manuel olarak belirlenmesi zaman alıcı ve çaba gerektiren önemli bir işlemdir. Görüntülerden eşleştirme yöntemiyle üretilen nokta bulutları, yoğun ve doğruluğu yüksek üç boyutlu (3B) konum bilgisi içermektedir. Binaların 3B nokta bulutlarından otomatik olarak çıkarılması geometrik düzensizlikleri, çıkarılacakları nokta bulutu yoğunluğu ve hassasiyeti açısından zor bir problemdir. Bu çalışmada, voksel temelli sekizdal (Octree) veri organizasyon metodu otomatikleştirilerek iyileştirilmiş Octree (I-Octree) yaklaşımı geliştirilmiş ve görüntülerden üretilen nokta bulutları üzerinde bina detaylarının otomatik çıkarımı ve düzgünleştirilmesi amaçlanmıştır. Elazığ bölgesinde seçilen çalışma alanında 3B nokta bulutu görüntülerden üretilmiş, zemin ve zemin üstü objeler SMRF metodu ile ayıklanmış, DBSCAN algoritması ile bina objeleri gürültülerden ayıklanarak sınıflandırılmış ve sekizdal ile I-Octree yöntemlerinin sınıflandırılan objelere uygulanması ile ortaya çıkarılan bina detaylarına ABORE metodu ile kenar düzgünleştirmesi işlemi uygulanmıştır. Otomatik olarak çıkarılan bina verileri çalışma alanını içeren 1/1000 ölçekli hâlihazır harita referans verisi desteğiyle piksel tabanlı tamlık (Cp), doğruluk (Cr), kalite (Q) ve F-skor (F-1) metrikleri ile doğrulanmıştır. Doğrulama sonuçları her bir metrik için maksimum değer olarak %94 üzerinde elde edilmiştir. Görüntülerden üretilmiş nokta bulutları üzerinden, geliştirilen I-Octree yaklaşımı ile bina detayı çıkarılması noktasında hızlı ve ucuz bir harita üretimi sürecine katkıda bulunabileceği sonucuna varılmıştır.

Anahtar Kelimeler

Fotogrametri, Nokta Bulutu, Bina Çıkarımı, I-Octree, Sayısallaştırma

Kaynakça

• Alidoost, F., Arefi, H. (2015). An image-based technique for 3d building reconstruction using multi-view UAV images. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives (C. 40).
• Alidoost, Fatemeh, Arefi, H., Tombari, F. (2019). 2D image-to-3D model: Knowledge-based 3D building reconstruction (3DBR) using single aerial images and convolutional neural networks (CNNs). Remote Sensing, 11(19).
• Bulatov, D., Häufel, G., Meidow, J., Pohl, M., Solbrig, P., Wernerus, P. (2014). Context-based automatic reconstruction and texturing of 3D urban terrain for quick-response tasks. ISPRS Journal of Photogrammetry and Remote Sensing, 93.
• Cao, Z., Fu, K., Lu, X., Diao, W., Sun, H., Yan, M., Yu, H., Sun, X. (2019). End-to-End DSM Fusion Networks for Semantic Segmentation in High-Resolution Aerial Images. IEEE Geoscience and Remote Sensing Letters, 16(11).
• Christian Rose, J., Paulus, S., Kuhlmann, H. (2015). Accuracy analysis of a multi-view stereo approach for phenotyping of tomato plants at the organ level. Sensors (Switzerland), 15(5).
• Dal Poz, A. P., Yano Ywata, M. S. (2020). Adaptive random sample consensus approach for segmentation of building roof in airborne laser scanning point cloud. International Journal of Remote Sensing, 41(6). Ester, M., Kriegel, H.-P., Sander, J., Xu, X. (1996). A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise. 2nd International Conference on Knowledge Discovery and Data Mining, 96(34).
• Gilani, S. A. N., Awrangjeb, M., Lu, G. (2016). An automatic building extraction and regularisation technique using LiDAR point cloud data and orthoimage. Remote Sensing, 8(3).
• Hermosilla, T., Ruiz, L. A., Recio, J. A., Estornell, J. (2011). Evaluation of automatic building detection approaches combining high resolution images and LiDAR data. Remote Sensing, 3(6).
• Jayaraj, P., Ramiya, A. M. (2018). 3D CityGML building modelling from lidar point cloud data. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives (C. 42).
• Ji, S., Wei, S., Lu, M. (2019). A scale robust convolutional neural network for automatic building extraction from aerial and satellite imagery. International Journal of Remote Sensing, 40(9).
• Karsli, F., Dihkan, M., Acar, H., Ozturk, A. (2016). Automatic building extraction from very high-resolution image and LiDAR data with SVM algorithm. Arabian Journal of Geosciences, 9(14).
• Kendall, A., Martirosyan, H., Dasgupta, S., Henry, P., Kennedy, R., Bachrach, A., Bry, A. (2017). End-to-End Learning of Geometry and Context for Deep Stereo Regression. Proceedings of the IEEE International Conference on Computer Vision (C. 2017-October).
• Lai, X., Yang, J., Li, Y., Wang, M. (2019). A building extraction approach based on the fusion of LiDAR point cloud and elevation map texture features. Remote Sensing, 11(14).
• Li, X., Ling, F., Foody, G. M., Du, Y. (2016). A superresolution land-cover change detection method using remotely sensed images with different spatial resolutions. IEEE Transactions on Geoscience and Remote Sensing, 54(7).
• Liu, J., Ji, S. (2020). A novel recurrent encoder-decoder structure for large-scale multi-view stereo reconstruction from an open aerial dataset. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition.
• Malihi, S., Valadan Zoej, M. J., Hahn, M., Mokhtarzade, M., Arefi, H. (2016). 3D Building Reconstruction Using Dense Photogrammetric Point Cloud. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLI-B3.
• Marullo, S., Patsaeva, S., Fiorani, L. (2018). Remote sensing of the coastal zone of the European seas. International Journal of Remote Sensing, 39(24).
• Nan, L., Wonka, P. (2017). PolyFit: Polygonal Surface Reconstruction from Point Clouds. Proceedings of the IEEE International Conference on Computer Vision (C. 2017-October).
• Ozdemir, E., Karsli, F., Kavzoglu, T., Bahadir, M., Yagmahan, A. (2021). A novel algorithm for regularization of building footprints using raw LiDAR point clouds. Geocarto International, 1-23.
• Rau, J. Y., Jhan, J. P., Hsu, Y. C. (2015). Analysis of oblique aerial images for land cover and point cloud classification in an Urban environment. IEEE Transactions on Geoscience and Remote Sensing, 53(3).
• Rutzinger, M., Rottensteiner, F., Pfeifer, N. (2009). A Comparison of Evaluation Techniques for Building Extraction from Airborne Laser Scanning. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2(1).
• Shao, Z., Yang, N., Xiao, X., Zhang, L., Peng, Z. (2016). A multi-view dense point cloud generation algorithm based on low-altitude remote sensing images. Remote Sensing, 8(5).
• Siddiqui, F. U., Teng, S. W., Awrangjeb, M., Lu, G. 2016. A robust gradient based method for building extraction from LiDAR and photogrammetric imagery. Sensors, 16(7).
• Sun, W., Wang, R. (2018). Fully Convolutional Networks for Semantic Segmentation of Very High Resolution Remotely Sensed Images Combined with DSM. IEEE Geoscience and Remote Sensing Letters, 15(3).
• Tran, T. N., Drab, K., Daszykowski, M. (2013). Revised DBSCAN algorithm to cluster data with dense adjacent clusters. Chemometrics and Intelligent Laboratory Systems, 120.
• Wu, Z., Wu, X., Zhang, X., Wang, S., Ju, L. (2019). Semantic stereo matching with pyramid cost volumes. Proceedings of the IEEE International Conference on Computer Vision (C. 2019-October).
• Yan, Y., Gao, F., Deng, S., Su, N. (2017). A hierarchical building segmentation in digital surface models for 3D reconstruction. Sensors, 17(2).
• Zhang, F., Prisacariu, V., Yang, R., Torr, P. H. S. (2019). GA-net: Guided aggregation net for end-to-end stereo matching. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (C. 2019-June).