IR2 stands for Instant Rain Rate, and is the technology behind the rainvolution. IR2 is a patented image processing algorithm, which gauges rain intensity in real time from pictures and videos shot during rain events*. IR2 lies on a simple assumption: since a photograph actually conveys 3D information, it is possible to count rain drops within a sampled volume, thus estimating rain intensity.
* It obviously returns null rainfall rates for images shot on sunny days. After all, this piece of information too may be useful. :-)
Let’s see how it works. Aperture, focal length and shutter speed determine the length of rain streaks and the depth of field of the image. After recognizing the streaks, IR2 uses these parameters to evaluate the actual dimension, position and velocity of every drop which falls within the focused volume detected in the photograph. This is the key information for computing instant rain rate and returning it to user’s devices.
Virtually every digital imaging device is able to provide useful images. Traditional surveillance cameras or dedicated webcams may be the nodes of a fixed network providing continuous rain monitoring in sites of particular interest. By processing sequences of images shot at high temporal frequency, IR2 returns robust and reliable rainfall rates per minute.
Smartphones, on the other hand, may be the nodes of a mobile network, whose data are more sporadic but still very precious to cover the most intense rain events, whenever and wherever they happen. By fusing data deriving from mobile devices and fixed nodes IR2 will be able to produce detail maps of rainfall patterns in real time.
IR2 is so fast, reliable and cheap that the spatial resolution of a IR2 monitoring network may be tenfold the resolution of a traditional rain gauge network, at the same overall cost. By connecting millions of cameras, IR2 will gather widespread real time data, which will flow in the greatest database of rainfall measurements in the world.