Measuring Water Levels by Handheld Smartphones – A contribution to exploit crowdsourcing in the spatio temporal densification of water gauging networks

Abstract

Global climate change leads to an increase in local heavy rainfall events causing nearly unpredictable flash floods worldwide. This paper introduces a novel and flexible low-cost water gauging technology, called Open Water Levels, using smartphones as low-cost measuring devices enabling the crowdsourcing of water levels on demand with accuracies of a few centimetres. This merely requires smartphone camera images of a riverbank and approximate values of the camera position and rotation measured by smartphone sensors. The images are analysed for the water line that is further projected into object space and intersected with a 3D model, e.g. from a GIS database, to derive water level information.

A Comparative Performance Analysis of Popular Deep Learning Models and Segment Anything Model (SAM) for River Water Segmentation in Close-Range Remote Sensing Imagery

“…validate numerical models or to acquire accurate data for early warning systems, the segmentation of river water in close-range Remote Sensing (RS) images captured by lowcost sensors becomes particularly significant [3], [4]. Indeed, close-range RS images captured by low-cost cameras (e.g., smartphone/surveillance cameras) are proven to facilitate the detection of subtle variations in river water properties and the surrounding terrain [3], [5], [6]. This presents, until yet, a rarely utilized and systematically investigated opportunity to extract nuanced insights into hydrological parameters or any related process (e.g., water level, water turbidity, floating debris, etc.)…”

“…This is mainly because most of these techniques depend on low-level/basic image features for object segmentation, which may not inherently capture complex spatial relationships, such as those found in river scenes in closerange images [8], [12]. In contrast, DL models, known for their automatic extraction of high-level semantic features from various data types, have recently offered more robust solutions in this specific aspect [5], [11], [12]. Furthermore, since DL models are typically trained on diverse datasets, they exhibit superior capabilities in handling sky reflections and structures on the water body, as depicted in Fig.…”