Revolutionizing Machine Vision with Praying Mantis-Inspired Technology
In an era where technology is rapidly advancing, the quest for superior machine vision has led researchers at the University of Virginia School of Engineering and Applied Science to look to the natural world for inspiration. Their groundbreaking work focuses on overcoming the significant limitations faced by existing visual systems in machines, including self-driving cars, which struggle with processing static or slow-moving objects in 3D space. What sets their approach apart is an unconventional model: the praying mantis.
Unlike many insects that have monocular vision, leading to poor depth perception, the praying mantis possesses a unique binocular vision, thanks to the overlapping fields of view between its left and right eyes. This biological marvel allows for detailed depth perception in three-dimensional space—an attribute the UVA team sought to replicate in artificial form.
By combining cutting-edge optoelectrical engineering with innovative edge computing, where data is processed close to where it is captured, the researchers have developed a new type of artificial compound eye. This biomimetic system mimics the praying mantis’s vision capabilities, promising a revolution in how machines perceive the world around them.
[INSERTIMAGE]“After an extensive study of the praying mantis’s eyes, we realized we needed to develop new technologies to emulate their biological capabilities,” explained Byungjoon Bae, a Ph.D. candidate leading the project. The team crafted artificial eyes that integrate microlenses with multiple photodiodes on a hemisphere, mimicking the mantis’s eye structure to achieve a wide field of view and superior depth perception.
These artificial eyes signal a breakthrough in machine vision technology, offering precise spatial awareness in real-time—a crucial component for dynamic applications ranging from autonomous vehicles and drones to surveillance systems and smart home devices.
One of the prototype’s most remarkable findings is its potential to reduce power consumption by over 400 times compared to conventional visual systems. This efficiency stems from the system’s ability to process visual information in real-time, thus nearly eliminating the delay and resources typically consumed by data transfer and external computation.
“The integration of flexible semiconductor materials, conformal devices, in-sensor memory, and unique algorithms represents a technological leap forward in our work,” Bae stated. This sophisticated approach allows the system to continuously monitor scene changes, identify pixel variations, and efficiently process spatial and motion data, imitating the way insects—and to a certain extent, humans—perceive their surroundings.
Special about the praying mantis’s vision, and now the UVA team’s artificial system, is the use of stereopsis, or seeing with both eyes to gauge depth, alongside the hemispherical compound eye design and motion parallax. This combination results in an unparalleled capacity for real-time, accurate 3D spatial perception.
“Our project not only demonstrates a significant scientific advancement,” said Kyusang Lee, the project’s advisor and an associate professor at UVA, “but also serves as an inspiring example of how biomimicry can provide ingenious solutions to complex engineering challenges.”
The team’s findings and the development of the stereoscopic artificial compound eyes were detailed in a recent paper published in the May 15 edition of Science Robotics, marking a pivotal step forward in the field of machine vision and edge computing technology. With support from the National Science Foundation and the U.S. Air Force Office of Scientific Research, this innovative work paves the way for advancements across a broad range of technological applications, mirroring the remarkable vision of nature’s own praying mantis.
