Next Article in Journal
Tree Crown Delineation Algorithm Based on a Convolutional Neural Network
Next Article in Special Issue
Building Extraction Using Orthophotos and Dense Point Cloud Derived from Visual Band Aerial Imagery Based on Machine Learning and Segmentation
Previous Article in Journal
Computer Vision and Deep Learning Techniques for the Analysis of Drone-Acquired Forest Images, a Transfer Learning Study
Article

Generalized Sparse Convolutional Neural Networks for Semantic Segmentation of Point Clouds Derived from Tri-Stereo Satellite Imagery

1
Research Institute for Computational Methods, Vienna University of Economics and Business, Welthandelsplatz 1, 1020 Vienna, Austria
2
Department of Geodesy and Geoinformation, Vienna University of Technology, Gußhausstraße 27-29, 1040 Vienna, Austria
3
Corporate Technology, Siemens AG Österreich, Siemensstraße 90, 1210 Vienna, Austria
4
Vermessung Schmid ZT GmbH, Büropark Donau, Inkustraße 1-7, Stiege 3, EG, 3400 Klosterneuburg, Austria
5
Department of Business and Management, Webster Vienna Private University, Praterstraße 23, 1020 Vienna, Austria
*
Authors to whom correspondence should be addressed.
Remote Sens. 2020, 12(8), 1289; https://doi.org/10.3390/rs12081289
Received: 6 March 2020 / Revised: 8 April 2020 / Accepted: 13 April 2020 / Published: 18 April 2020
(This article belongs to the Special Issue Machine and Deep Learning for Earth Observation Data Analysis)
We studied the applicability of point clouds derived from tri-stereo satellite imagery for semantic segmentation for generalized sparse convolutional neural networks by the example of an Austrian study area. We examined, in particular, if the distorted geometric information, in addition to color, influences the performance of segmenting clutter, roads, buildings, trees, and vehicles. In this regard, we trained a fully convolutional neural network that uses generalized sparse convolution one time solely on 3D geometric information (i.e., 3D point cloud derived by dense image matching), and twice on 3D geometric as well as color information. In the first experiment, we did not use class weights, whereas in the second we did. We compared the results with a fully convolutional neural network that was trained on a 2D orthophoto, and a decision tree that was once trained on hand-crafted 3D geometric features, and once trained on hand-crafted 3D geometric as well as color features. The decision tree using hand-crafted features has been successfully applied to aerial laser scanning data in the literature. Hence, we compared our main interest of study, a representation learning technique, with another representation learning technique, and a non-representation learning technique. Our study area is located in Waldviertel, a region in Lower Austria. The territory is a hilly region covered mainly by forests, agriculture, and grasslands. Our classes of interest are heavily unbalanced. However, we did not use any data augmentation techniques to counter overfitting. For our study area, we reported that geometric and color information only improves the performance of the Generalized Sparse Convolutional Neural Network (GSCNN) on the dominant class, which leads to a higher overall performance in our case. We also found that training the network with median class weighting partially reverts the effects of adding color. The network also started to learn the classes with lower occurrences. The fully convolutional neural network that was trained on the 2D orthophoto generally outperforms the other two with a kappa score of over 90% and an average per class accuracy of 61%. However, the decision tree trained on colors and hand-crafted geometric features has a 2% higher accuracy for roads. View Full-Text
Keywords: 3D segmentation; deep learning; derived point clouds; tri-stereo; Very High Resolution (VHR) Satellite Imagery; 2.5D segmentation; image segmentation; semantic segmentation; machine learning 3D segmentation; deep learning; derived point clouds; tri-stereo; Very High Resolution (VHR) Satellite Imagery; 2.5D segmentation; image segmentation; semantic segmentation; machine learning
Show Figures

Graphical abstract

MDPI and ACS Style

Bachhofner, S.; Loghin, A.-M.; Otepka, J.; Pfeifer, N.; Hornacek, M.; Siposova, A.; Schmidinger, N.; Hornik, K.; Schiller, N.; Kähler, O.; Hochreiter, R. Generalized Sparse Convolutional Neural Networks for Semantic Segmentation of Point Clouds Derived from Tri-Stereo Satellite Imagery. Remote Sens. 2020, 12, 1289. https://doi.org/10.3390/rs12081289

AMA Style

Bachhofner S, Loghin A-M, Otepka J, Pfeifer N, Hornacek M, Siposova A, Schmidinger N, Hornik K, Schiller N, Kähler O, Hochreiter R. Generalized Sparse Convolutional Neural Networks for Semantic Segmentation of Point Clouds Derived from Tri-Stereo Satellite Imagery. Remote Sensing. 2020; 12(8):1289. https://doi.org/10.3390/rs12081289

Chicago/Turabian Style

Bachhofner, Stefan, Ana-Maria Loghin, Johannes Otepka, Norbert Pfeifer, Michael Hornacek, Andrea Siposova, Niklas Schmidinger, Kurt Hornik, Nikolaus Schiller, Olaf Kähler, and Ronald Hochreiter. 2020. "Generalized Sparse Convolutional Neural Networks for Semantic Segmentation of Point Clouds Derived from Tri-Stereo Satellite Imagery" Remote Sensing 12, no. 8: 1289. https://doi.org/10.3390/rs12081289

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop