The video should autostart with english captions activated.
This multispectral composite image in coded false colour information showing a local game reserve of the project was awarded a “distinction” at a competition for scientific photography of the Swiss National Science Foundation.
The hidden forest – Award in the category “Object of study”
Adrian Meyer (Scientific collaborator, University of Applied Sciences Northwestern Switzerland – FHNW)
Description: This composite false-colour map was generated with a drone-mounted multispectral sensor flying over the Southern Black Forest near Bad Säckingen, Germany, North of Basel. The use of near-infrared light make visible the plants’ metabolic state and health, which change their reflectance patterns. This hidden information would remain invisible to the human eye if science would not keep pushing the use of advanced modern technology. The complex spectral colour codes can be analysed by artificial intelligence algorithms to make predictions of the locations of local wildlife populations. Details: Airinov Multispec 4C sensor mounted on a eBee research drone built by EPFL spin-off SenseFly; the orthoprojected mosaic was processed from 431 individual multispectral images taken from an altitude of 100 metres above ground level at a ground resolution of 10 cm per pixel.
Jury comment: A picture which, at first glance, might be considered beautiful, shows how science finds new ways to present altered states of plants’ and tree’ metabolic health. As opposed to the often-used flat colour coding, this strongly contrasted explosion of colours offers a vibrant, but at the same time confusing visual investigation of nature. The outer boundary adds a puzzling element that enables the viewer to adopt a dazzling, quasi drone-like aerial view of phenomena that, indeed, evade human perception.
CC BY-NC-ND 2.0
Today‘s @PyBasel Meetup in the new #FHNW Campus Muttenz starts with two @igeoFHNW presentations: @stefan_hochuli on #DeepLearning for the Swiss land cover classification and Adrian Meyer on #UAV-based wildlife monitoring using #DeepLearning. pic.twitter.com/Hoa7dW2Xt6
— Stephan Nebiker (@snebiker) September 6, 2018