Dataset: Image-based water stage measurement with deep learning

Other contributing persons, institutions or organisations	eu - Funder
Person(s) who is (are) responsible for the content of the research data	Eltner, Anette
Abstract	This dataset contains the results of the application of deep learning to classify water areas in images captured with a low-cost Raspberry Pi camera. Furthermore, the photogrammetrically measured 3D point cloud of the shore and the extracted water stage information is included as well as the comparison between camera gauge measurement and an independent reference gauge measurement.
Year or period of data production	2020
Publication year	2020
Publisher	Technische Universität Dresden
Content of the research data	Image, Dataset: extracted water contours, corresponding calculated water stages, 3D point cloud of the shore as well as comparison to reference gauge measurements images with classified water area
Holder of usage rights	Technische Universität Dresden
Usage rights of the data	CC-BY-4.0
Discipline(s)	Geography	de
Title of the dataset	Dataset: Image-based water stage measurement with deep learning

Image-based water stage measurement using deep learning [1]
A workflow is introduced to automatically measure water stages based on image measurements using deep learning. So far, most camera gauges do not provide the needed robustness to achieve accurate water stage measurements because of changing environmental conditions. The novel, suggested approach is based on two CNNs (i.e. FCN and SegNet) to identify water in imagery. The image information is transformed into metric water level values intersecting the extracted water contour with a 3D model. The workflow allows for the densification of river monitoring networks based on low-cost camera gauges in various scenarios.