Dragonflies have long been known to be some of the best when it comes to tracking their prey. While they may not have the largest brain or the best vision among predatory insects, they can chase their target at speeds of up to 60 miles per hour and their capture rate is around 97 percent successful.
This unique dragon ability is what a team of scientists at Australia's University of Adelaide are trying to replicate using a new algorithm they developed. Their goal is to enhance the visual capabilities of robots through the use of the program.
Steven Wiederman, a neuroscientist at Adelaide and one of the co-authors of the study, said that the performance of the algorithm can provide real-time applications using simple computer processors.
The researchers were inspired to develop the algorithm after analyzing the responses of neurons in a dragonfly brain to visual stimuli.
They have tested the new program in visual settings and are now getting ready to use it on a bio-inspired autonomous robot to determine its practical benefits.
"Instead of just trying to keep the target perfectly centered on its field of view, our system locks on to the background and lets the target move against it," Zahra Bagheri, a PhD student at Adelaide and the lead author of the research, said.
"This reduces distractions from the background and gives time for underlying brain-like motion processing to work. It then makes small movements of its gaze and rotates towards the target to keep the target roughly frontal."
The algorithm is designed to allow robots to pursue their target without getting sidetracked by distractions in their surroundings.
Bagheri explained the difficulty of detecting and tracking small objects in complicated environments.
"Consider a cricket or baseball player trying to take a match-winning catch in the outfield. They have seconds or less to spot the ball, track it and predict its path as it comes down against the brightly colored backdrop of excited fans in the crowd - all while running or even diving towards the point where they predict it will fall."
Experts believe that the dragonfly-inspired algorithm will help robot developers combine accurate vision with flexible muscles and fast reflexes in their designs.
This new technology can be applied to improve the abilities of robots used for disaster response and rescue efforts such as the ones featured in the Defense Advanced Research Projects Agency (DARPA) Robotics Challenge.
The algorithm will allow rescue robots to quickly search and locate people who need saving during a catastrophe.
The University of Adelaide study is published in the Journal of The Royal Society Interface.
Photo: Roberto Jorge | Flickr
