With 'robotic cat's eyes', drones can track targets more precisely than ever before.
A new computer vision system based on a cat's eye design could enable future drones and other military robots to track targets in low-visibility and dynamic environments. Such robotic “eyes” will allow drones to see the surrounding world more accurately than ever before.
LiveScience writes about this.
Robots, drones, self-driving cars and other autonomous systems are becoming increasingly common, but they still “see” poorly in any environment and under any conditions. For example, self-driving cars don't work well in rain or fog because these conditions make it difficult for the car's sensors and cameras to work.
Now scientists have developed a new vision system that uses advanced lenses and sensors modeled after the structure of a cat's eye, providing improved detection and recognition of objects.
They chose the cat's eye as a prototype because these furry animals can see perfectly in both light and darkness. During the day, the cat's pupil — it's a vertical slit-like shape that filters light and reduces glare, helping the cat focus. In twilight and darkness, the pupil dilates to let in more light — using a reflective layer called the tapetum lucidum, which improves vision by reflecting visible light back through the retina and amplifying the rays available to the photoreceptors.
200% Deposit Bonus up to €3,000 180% First Deposit Bonus up to $20,000Similarly, the new system includes a slit diaphragm that helps filter out unwanted light and highlight key objects in bright conditions, then uses reflective layers similar to those found in cats' eyes to improve visibility in low-light conditions.
“Robotic cameras often struggle to spot objects against busy or camouflaged backgrounds, especially when lighting conditions change. Our design solves this problem by allowing robots to blur out unnecessary details and focus on important objects,” — said the study's lead author Young Min Song, a professor of electronic engineering at the Gwangju Institute of Science and Technology (GIST) in South Korea.
The scientist emphasized that the computer vision system relies on specialized lenses rather than powerful computer processing, so it is more energy efficient.
When the researchers tested their system, they found that it successfully blurs background objects. objects and at the same time keeps the focus on the target object. They also used a neural network — a set of machine learning algorithms designed to process information in a similar way to the human brain — to help the system better perceive important objects.
According to the researchers, to deploy the system in practical and commercial conditions, they will need to improve the pixel resolution of the field of view. But in the future, it could be integrated into robots and other machines, including military drones and surveillance robots. Scientists say the technology will soon be sophisticated enough to allow robots to detect, track and recognize targets in ever-changing environments where they can be very difficult to see with conventional cameras.