Emmanuel P. Baltsavias

TL;DR: In this article, a method for lake ice monitoring using low spatial resolution (250m-1000m) satellite images to determine whether a lake is frozen or not is described, where the most useful channels to solve the problem are selected with xgboost and visual analysis of histograms of reference data, while the classification is done with nonlinear support vector machine (SVM).

TL;DR: This work serves as a pilot project to illustrate the capabilities of digital photogrammetry and geographic information system techniques and also aims at generating some expertise for a longer-term objective of a national project that can be carried out in cooperation with the Department of Antiquities, Jordan.

TL;DR: In this paper, the authors present a lake ice monitoring system based on the automatic analysis of Sentinel-1 Synthetic Aperture Radar (SAR) data with a deep neural network, which reaches mean intersectionover-union (mIoU) scores >90% on average, and >84% even for the most difficult lake.

TL;DR: In this article, the authors report on progress towards the automated generation of geospatial databases of informal settlements from large scale, low-cost, digital still-video imagery, focusing on shack extraction as the predominant data requirement.

TL;DR: This work presents a lake ice monitoring system based on the automatic analysis of Sentinel-1 Synthetic Aperture Radar data with a deep neural network, and casts ice detection as a two class (frozen, non-frozen) semantic segmentation problem and solves it using a state-of-the-art deep convolutional network (CNN).