- C. Persello (1)
- Claudio Persello (1)
- M. Sommer (1)
- A.J.A.M. Temme (1)
- D. Tuia (1)
- Devis Tuia (1)
- G. Vosselman (1)
- George Vosselman (1)
- M.Y. Yang (1)
- Michael Ying Yang (1)
- Z. Zhang (1)
- Zhenchao Zhang (1)
Detecting Building Changes between Airborne Laser Scanning and Photogrammetric Data
Zhang, Zhenchao ; Vosselman, George ; Gerke, Markus ; Persello, Claudio ; Tuia, Devis ; Yang, Michael Ying - \ 2019
Remote Sensing 11 (2019)20. - ISSN 2072-4292 - 17 p.
Detecting topographic changes in an urban environment and keeping city-level point clouds up-to-date are important tasks for urban planning and monitoring. In practice, remote sensing data are often available only in different modalities for two epochs. Change detection between airborne laser scanning data and photogrammetric data is challenging due to the multi-modality of the input data and dense matching errors. This paper proposes a method to detect building changes between multimodal acquisitions. The multimodal inputs are converted and fed into a light-weighted pseudo-Siamese convolutional neural network (PSI-CNN) for change detection. Different network configurations and fusion strategies are compared. Our experiments on a large urban data set demonstrate the effectiveness of the proposed method. Our change map achieves a recall rate of 86.17%, a precision rate of 68.16%, and an F1-score of 76.13%. The comparison between Siamese architecture and feed-forward architecture brings many interesting findings and suggestions to the design of networks for multimodal data processing.
Change detection between digital surface models from airborne laser scanning and dense matching using convolutional neural networks
Zhang, Z. ; Vosselman, G. ; Gerke, M. ; Persello, C. ; Tuia, D. ; Yang, M.Y. - \ 2019
In: ISPRS Geospatial Week 2019, 10–14 June 2019, Enschede, The Netherlands. - ISPRS (ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences ) - p. 453 - 460.
Airborne Laser Scanning - Change Detection - Convolutional Neural Network (CNN) - Dense Image Matching - Digital Surface Model (DSM)
Airborne photogrammetry and airborne laser scanning are two commonly used technologies used for topographical data acquisition at the city level. Change detection between airborne laser scanning data and photogrammetric data is challenging since the two point clouds show different characteristics. After comparing the two types of point clouds, this paper proposes a feed-forward Convolutional Neural Network (CNN) to detect building changes between them. The motivation from an application point of view is that the multimodal point clouds might be available for different epochs. Our method contains three steps: First, the point clouds and orthoimages are converted to raster images. Second, square patches are cropped from raster images and then fed into CNN for change detection. Finally, the original change map is post-processed with a simple connected component analysis. Experimental results show that the patch-based recall rate reaches 0.8146 and the precision rate reaches 0.7632. Object-based evaluation shows that 74 out of 86 building changes are correctly detected.
On the role of hydrologic processes in soil and landscape evolution modeling : concepts, complications and partial solutions
Meij, W.M. van der; Temme, A.J.A.M. ; Lin, H.S. ; Gerke, H.H. ; Sommer, M. - \ 2018
Earth-Science Reviews 185 (2018). - ISSN 0012-8252 - p. 1088 - 1106.
The ability of water to transport and transform soil materials is one of the main drivers of soil and landscape development. In turn, soil and landscape properties determine how water is distributed in soil landscapes. Understanding the complex dynamics of this co-evolution of soils, landscapes and the hydrological system is fundamental in adapting land management to changes in climate. Soil-Landscape Evolution Models (SLEMs) are used to simulate the development and evolution of soils and landscapes. However, many hydrologic processes, such as preferential flow and subsurface lateral flow, are currently absent in these models. This limits the applicability of SLEMs to improve our understanding of feedbacks in the hydro-pedo-geomorphological system. Implementation of these hydrologic processes in SLEMs faces several complications related to calculation demands, limited methods for linking pedogenic and hydrologic processes, and limited data on quantification of changes in the hydrological system over time. In this contribution, we first briefly review processes and feedbacks in soil-landscape-hydrological systems. Next, we elaborate on the development required to include these processes in SLEMs. We discuss the state-of-the-art knowledge, identify complications, give partial solutions and suggest important future development. The main requirements for incorporating hydrologic processes in SLEMs are: (1) designing a model framework that can deal with varying timescales for different sets of processes, (2) developing and implementing methods for simulating pedogenesis as a function of water flow, (3) improving and implementing knowledge on the evolution and dynamics of soil hydraulic properties over different timescales, and (4) improving the database on temporal changes and dynamics of flow paths.