Missing climbers and hikers are common. Rapid rescue is often a matter of life and death, especially if they have been in an accident or are at risk of extreme temperatures.
In such cases, helicopters with thermal cameras that detect a person’s thermal footprint are often used. At least in theory. In fact, on hot days, treetops and other places can reach body temperature, making it impossible to distinguish between lost people. One of his ways of optimizing tracking is to apply artificial intelligence and machine learning to interpret the acquired images. That’s the suggestion of three researchers at the Johannes Kepler University in Austria. The results were just published in the scientific journal Nature Machine Intelligence.
System developers were inspired by clustered radio telescopes. This will combine multiple images into one image of her. The first step in the artificial intelligence software used is to generate a single image obtained from a photograph of an infrared camera mounted on a drone or helicopter. The image is then processed to increase the depth of field and distinguish between the ground and the canopy. This is achieved by calculating different focal lengths. Additionally, the software used a library of images to train machine learning functions. To demonstrate its effectiveness in the real world, they used a group of volunteers who went into the forest. Tests have shown that their system is 87% to 95% effective against him, as opposed to the 25% success rate of traditional infrared cameras.
Drone-mounted environment sensor
Besides locating lost people, drones can also be applied to forest surveillance. One current trend is to install autonomous sensors in forested areas, some powered by triboelectricity. The problem is that it often has to be manually installed in an inaccessible location. The Aeronautical Robotics Laboratory at Imperial College London is now looking to one of the new applications for drones to speed up the process.
The new system allows drones to shoot wireless sensors in the form of darts at tree trunks and branches. Drones are equipped with cameras to identify suitable locations. The sensor also incorporates a smart material that changes shape when heated and adheres to surfaces. Once installed, it provides valuable information about the ecosystem. On the one hand, it provides data on temperature, light, humidity or animal movements. It also helps prevent fires by detecting risk situations caused by low humidity and high temperatures and the first signs of fire.
For now, the drones used will need an operator, but the medium-term goal is to have autonomous drones that can go deep into the forest and find the best places to place sensors. . In the long term, it will be possible to build sensor networks in areas such as the Amazon jungle, where dense vegetation makes it very difficult to operate.
These environmental protection technologies join other very interesting initiatives such as artificial intelligence-powered hot air balloons and satellite systems monitoring the health of the oceans.
sauce: TechXplore, TechXplore
