 |
A group of researchers at Johannes Kepler University has used artificial intelligence(AI) to enhance thermal imaging camera searches of individuals lost in the woods or Forests. In their paper published within the journal Nature Machine Intelligence, David Schedl, Indrajit Kurmi, and Oliver Bimber describe how they utilized a very deep learning network to the issue of individuals lost within the woods and how nicely it worked.
When individuals lost in woods, search and rescue specialists use helicopters to fly over the realm of the place they're most likely to be found. In addition to simply scanning the bottom ground below, the researchers use binoculars and thermal imaging cameras.
It's hoped that such cameras will spot and highlight differences in body temperature of folks on the ground versus their environment making them easier to identify. Sadly, in some situations, thermal imaging doesn't work as intended due to vegetation covering subsoil or the solar heating of the trees to a temperature that's just like the body temperature of the person who is lost.
In this new tech effort, the researchers sought to beat these problems through the use of a deep learning utility to enhance the images that are made.
The solution the group developed involved using an AI application to process a number of images of a given area. They examine it by utilizing Artificial Intelligence to process information from several radio telescopes. Doing so permits multiple telescopes to function as a single giant telescope. In like method, the AI application they used allowed a number of thermal images taken from a helicopter (or drone) to create a picture as if it were captured by a camera with a lot larger lens.
After processing, the pictures that were produced had a lot higher depth of area; in them, the tops of the trees at wood appeared blurred while individuals on the ground became much more recognizable.
When individuals lost in woods, search and rescue specialists use helicopters to fly over the realm of the place they're most likely to be found. In addition to simply scanning the bottom ground below, the researchers use binoculars and thermal imaging cameras.
It's hoped that such cameras will spot and highlight differences in body temperature of folks on the ground versus their environment making them easier to identify. Sadly, in some situations, thermal imaging doesn't work as intended due to vegetation covering subsoil or the solar heating of the trees to a temperature that's just like the body temperature of the person who is lost.
In this new tech effort, the researchers sought to beat these problems through the use of a deep learning utility to enhance the images that are made.
The solution the group developed involved using an AI application to process a number of images of a given area. They examine it by utilizing Artificial Intelligence to process information from several radio telescopes. Doing so permits multiple telescopes to function as a single giant telescope. In like method, the AI application they used allowed a number of thermal images taken from a helicopter (or drone) to create a picture as if it were captured by a camera with a lot larger lens.
After processing, the pictures that were produced had a lot higher depth of area; in them, the tops of the trees at wood appeared blurred while individuals on the ground became much more recognizable.
To train the AI system: the researchers needed to create their own database of photos. They used drones to take footage of volunteers on the ground in all kinds of positions.
Testing of the system confirmed it to be roughly 87 to 95 correct compared to just 25 % accurate for traditional thermal images. The researchers recommend their system is prepared for use by search and rescue crews and may be utilized by law enforcement, the military, or wildlife management teams.
Testing of the system confirmed it to be roughly 87 to 95 correct compared to just 25 % accurate for traditional thermal images. The researchers recommend their system is prepared for use by search and rescue crews and may be utilized by law enforcement, the military, or wildlife management teams.
[Image and Information are taken from Techxplore.com]
Post A Comment:
0 comments: