rain-floodIt seems that a new kind of machine-learning algorithm that works in tandem with social media and remote sensing, has managed to successfully identify flooded areas, all without the need to fall back on the use of satellites. This is definitely a breakthrough that is worth checking out, as it makes use of a combination of Twitter and Flickr, in addition to remote sensor data, so that flooded areas are identified – at least this is what a bunch of university researchers claim. They also cite that this is a far speedier method as opposed to using publicly available satellite images, as such imaging could take days before being made available.

It is the algorithms used which are the “glue” in holding everything together, and the computer will then pick up as to what is and what is not water in the event of a flood. This is made possible by analyzing publicly posted images as well as going through the numerous public tweets and posts which are generated throughout an incident – especially in urban flooding situations. This makes satellite analysis secondary – and perhaps redundant. Of course, when one talks about this in the context of the interiors or places without any good telecommunications infrastructure in the first place, satellite analysis would still play a major role, but at least there is an alternative to explore now.

Filed in Computers. Read more about . Source: networkworld

Discover more from Ubergizmo

Subscribe now to keep reading and get access to the full archive.

Continue reading