A team of scientists from the University of Virginia School of Engineering and Applied Science has developed Innovative biomimetic vision system Inspired by the unique visual possibilities praying Mantis eyes. This innovation aims to increase the efficiency of various technologies, including self -propelled cars, UAV and robotic assembly lines, while dealing with a significant challenge in AI -based systems: the inability to accurately perceive static or slow facilities in 3D space.
For example, independent cars currently rely on visual systems, which, like the complex eyes of most insects, stand out, tracking and offer a wide field of view, but fight depth perception. However, praying mantis stands out as an exception. His eyes that overlap in their field of view ensure binocular vision – allowing him to perceive the depth in 3D space, which is a critical ability that the research team tried to repeat.
Scientists, led by the doctor, Byungjoon Bae, designed artificial complex eyes that imitate this biological ability. These “eyes” integrate microlens and many photodiodes using flexible semiconductor materials that imitate convex shapes and positions faceted in the eyes of Mantis. This design allows for a wide field of view while maintaining exceptional depth perception.
According to Bae, their system provides spatial awareness in real time, which is crucial for applications that affect dynamic environments. One of the key innovations in this system is the use of edge calculations – data processing directly in or nearby sensors. This approach significantly reduces data processing time and energy consumption, reaching over 400 times a reduction in energy consumption compared to traditional visual systems. This makes technology particularly suitable for low -power vehicles, drones, robotic systems and intelligent home appliances.
The team's work shows how these artificial complex eyes can constantly monitor the changes in the stage, identifying and encoding which pixels have changed. This method reflects the way insects processes visual information, using traffic parallax to distinguish between close and distant objects, and to perceive movement and spatial data.
By combining advanced materials, innovative algorithms and a deep understanding of biological vision systems, scientists have created a computer vision system that can revolutionize AI applications. This biomimetic approach not only increases the accuracy and efficiency of visual processing, but also opens up new possibilities of the future of AI technology.
As self -propelled cars, UAV and other AI systems, the integration of such biomimetic vision systems can mean a serious jump, making these technologies safer and more reliable in real environments.
