Advancing Opportunistic Sensing in Hydrology: A Novel Approach to Measuring Rainfall With Ordinary Surveillance Cameras

Abstract

“Opportunistic sensing” represents an appealing idea for collecting unconventional data with broad spatial coverage and high resolution, but few studies have explored its feasibility in hydrology. This study develops a novel approach to measuring rainfall intensity in real‐world conditions based on videos acquired by ordinary surveillance cameras. The proposed approach employs a convex optimization algorithm to effectively decompose a rainy image into two layers: a pure rain‐streak layer and a rain‐free background layer, where the rain streaks represent the motion blur of falling raindrops. Then, it estimates the instantaneous rainfall intensity via geometrical optics and photographic analyses. We investigated the effectiveness and robustness of our approach through synthetic numerical experiments and field tests. The major findings are as follows. First, the decomposition‐based identification algorithm can effectively recognize rain streaks from complex backgrounds with many disturbances. Compared to existing algorithms that consider only the temporal changes in grayscale between frames, the new algorithm successfully prevents false identifications by considering the intrinsic visual properties of rain streaks. Second, the proposed approach demonstrates satisfactory estimation accuracy and is robust across a wide range of rainfall intensities. The proposed approach has a mean absolute percentage error of 21.8%, which is significantly lower than those of existing approaches reported in the literature even though our approach was applied to a more complicated scene acquired using a lower‐quality device. Overall, the proposed low‐cost, high‐accuracy approach to vision‐based rain gauging significantly enhances the possibility of using existing surveillance camera networks to perform opportunistic hydrology sensing.

Identifying changing precipitation extremes in Sub-Saharan Africa with gauge and satellite products

“…In the field of computer vision, R×1day automated detection of hydrologically relevant spatial features from complex data is improving. This type of information has been used in machine learning systems for hydrologic prediction (Jiang et al 2018) and in opportunistic sensing projects such as for measuring rainfall intensity from ordinary surveillance cameras (Jiang et al 2019). The relationship between large-scale variability, anthropogenic forcing, and precipitation extremes requires more study.…”

Balancing Open Science and Data Privacy in the Water Sciences

“…High-resolution spatial data include satellite data (and derived products), outputs of hydrological models, and other geospatial datasets. Geospatial data are commonly used in the hydrologic sciences, and unmanned aerial vehicles (i.e., drones; Kelleher et al, 2018), traffic/surveillance cameras (Leitão et al, 2018;Jiang et al, 2019), and increasing access to satellite data are likely to make these data less costly to collect and more widely available. Despite not meeting traditional definitions of human subject research, this type of data could be sensitive at the individual and community levels (Rissman et al, 2017).…”

Rainfall monitoring with opportunistic sensors

“…Wolf et al, 2017), the voltage induced in a sensing coil upon impact (impact, e.g. Joss & Waldvogel, 1967), 2D video recordings of falling droplets (light-sheet, e.g. Kruger & Krajewski, 2002) or the length and size of the interference in an laser beam created by falling hydrometeors (optical, e.g.…”

“…Weather observations can be human-based, like citizen weather observations provided via smartphone apps (Elmore et al, 2014;Guo et al, 2019), twitter (De Vasconcelos et al, 2016) and volunteers actively reporting daily rainfall at their location (Cifelli et al, 2005;Illingworth et al, 2014;Reges et al, 2016). Recent years have seen various types of rainfall information retrieved from technology as well, like readings from windshield wipers and optical sensors in cars (Rabiei et al, 2013) and rainfall intensity from camera images (Allamano et al, 2015;Jiang et al, 2019). Reviews on crowdcourcing atmospheric data in geophysics are provided by Muller et al (2015) and Zheng et al (2018).…”

“…This is problematic as high-resolution (in space and time) flood forecasting requires high-resolution precipitation input in order to produce meaningful results, especially in urban areas (Berne et al, 2004;Emmanuel et al, 2012;Ochoa-Rodriguez et al, 2015;Cristiano et al, 2017). Crowdsourcing has been investigated as a strategy to obtain more rainfall observations, ranging from studies exploring citizen observations collected via smartphone apps (Elmore et al, 2014;Guo et al, 2019), active daily rainfall amounts reported by volunteers (Cifelli et al, 2005;Illingworth et al, 2014;Reges et al, 2016), rainfall intensities from camera images (Allamano et al, 2015;Jiang et al, 2019), rain intensity and occurrence from car sensors (Rabiei et al, 2013), derived weather information from twitter messages (De Vasconcelos et al, 2016) to simulation studies incorporating those techniques (e.g. Mazzoleni et al, 2017;Yang & Ng, 2017).…”