IM2GPS: estimating geographic information from a single image
James Hays, Alexei A. Efros
Estimating geographic information from an image is an excellent, difficult high-level computer vision problem whose time has come. The emergence of vast amounts of geographically-calibrated image data is a great reason for computer vision to start looking globally - on the scale of the entire planet! In this paper, we propose a simple algorithm for estimating a distribution over geographic locations from a single image using a purely data-driven scene matching approach. For this task, we leverage a dataset of over 6 million GPS-tagged images from the Internet. We represent the estimated image location as a probability distribution over the Earthpsilas surface. We quantitatively evaluate our approach in several geolocation tasks and demonstrate encouraging performance (up to 30 times better than chance). We show that geolocation estimates can provide the basis for numerous other image understanding tasks such as population density estimation, land cover estimation or urban/rural classification.
1033 sitasi
en
Computer Science, Geography
Geographical landslide early warning systems
F. Guzzetti, S. L. Gariano, S. Peruccacci
et al.
Abstract The design, implementation, management, and verification of landslide early warning systems (LEWSs) are gaining increasing attention in the literature and among government officials, decision makers, and the public. Based on a critical analysis of nine main assumptions that form the rationale for landslide forecasting and early warning, we examine 26 regional, national, and global LEWSs worldwide from 1977 to August 2019. We find that currently only five nations, 13 regions, and four metropolitan areas benefit from LEWSs, while many areas with numerous fatal landslides, where landslide risk to the population is high, lack LEWSs. Operational LEWSs use information from rain gauge networks, meteorological models, weather radars, and satellite estimates; and most systems use two sources of rainfall information. LEWSs use one or more types of landslide forecast models, including rainfall thresholds, distributed slope stability models, and soil water balance models; and most systems use landslide susceptibility zonations. Most LEWSs have undergone some form of verification, but there is no accepted standard to check the performance and forecasting skills of a LEWS. Based on our review, and our experience in the design, implementation, management, and verification of geographical LEWSs in Italy, we conclude that operational forecast of weather-induced landslides is feasible, and it can help reduce landslide risk. We propose 30 recommendations to further develop and improve geographical LEWSs, and to increase their reliability and credibility. We encourage landslide forecasters and LEWSs managers to propose open standards for geographical LEWSs, and we expect our work to contribute to this endeavour.
403 sitasi
en
Environmental Science