Abstract
Abstract. This paper explores the role deep convolutional neural networks play in automated extraction of linear structures using semantic segmentation techniques in Digital Terrain Models (DTMs). DTM is a regularly gridded raster created from laser scanning point clouds and represents elevations of the bare earth surface with respect to a reference. Recent advances in Deep Learning (DL) have made it possible to explore the use of semantic segmentation for detection of terrain structures in DTMs. This research examines two novel and practical deep convolutional neural network architectures i.e. an encoder-decoder network named as SegNet and the recent state-of-the-art high-resolution network (HRNet). This paper initially focuses on the pixel-wise binary classification in order to validate the applicability of the proposed approaches. The networks are trained to distinguish between points belonging to linear structures and those belonging to background. In the second step, multi-class segmentation is carried out on the same DTM dataset. The model is trained to not only detect a linear feature, but also to categorize it as one of the classes: hollow ways, roads, forest paths, historical paths, and streams. Results of the experiment in addition to the quantitative and qualitative analysis show the applicability of deep neural networks for detection of terrain structures in DTMs. From the deep learning models utilized, HRNet gives better results.
Highlights
The extraction of information, including linear structures, plays an important role in a wide range of disciplines where topographic features are used in spatial analysis, e.g. hydrological applications and archaeological applications
Digital Elevation Model (DEM) data Kazimi et al (Kazimi et al, 2020) recently proposed a Multi-Modal High Resolution network named MM-HR which is based on high-resolution network (HRNet) (Ke et al, 2019) and multi-modal deep learning approach (MM) (Du et al, 2019). Their proposed architecture with fewer parameter outperformed the MM architecture on the dataset of archaeological mining structures from Harz. In this project, based on literature analysis and our preliminary experiments with an intent to contribute to the field of remote sensing following recent works in (Kazimi et al, 2019b, Kazimi et al, 2020), we explore the use of semantic segmentation by doing experiments using two different Convolutional Neural Networks (CNNs) architectures, an encoderdecoder network named SegNet (Badrinarayanan et al, 2017a), and a high resolution network called HRNet (Sun et al, 2019) for the sole purpose of extraction of linear structures in Digital Terrain Models (DTMs)
As discussed in the previous sections of this project, the goal of binary-classification task is to detect the linear features from DTM data using both traditional and Deep Learning (DL) methods
Summary
The extraction of information, including linear structures, plays an important role in a wide range of disciplines where topographic features are used in spatial analysis, e.g. hydrological applications and archaeological applications. Methods used for line extraction from DTMs can be divided into two categories: Physical water flow simulation-based methods (O’Callaghan and Mark, 1984, Jenson and Domingue, 1988, Quinn et al, 1991, Tarboton, 1997) and geometrical morphological analysis-based methods (Chang et al, 1998, Gülgen and Gökgöz, 2004, Zhang et al, 2013, Zou and Weng, 2017, Peucker and Douglas, 1975) as mentioned in (Tsai, 2019) The former conduct running water simulation on terrain surface and the latter, geometric approach identify feature candidates for extracting terrain feature lines
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
