New Vision Technology thanks to Locusts

Researchers from the University of Lincoln and Newcastle University in the UK designed a new vision technology for self-moving robots by scrutinizing locusts and their visual ability.

This is a 4 year research project which began in 2011. The project is funded by the Seventh Framework Program of the European Union. Professor Shigang Yue from the School of Computer Science at the University of Lincoln and Doctor Claire Rind from the Institute of Neuroscience at Newcastle University lead the project.  Researchers from the University of Hamburg in Germany, Tsinghua University and Xi’an Jiaotong University in China are also in collaboration with them.

The main objective of the research is to create an international platform for visual systems that are biologically inspired. The aim is to make a new vision technology available to different applications. The system based on visual input enables a robot to find its way and recognize the objects around in dynamic environments. Taking a locust as a model is beneficial since this collision avoidance system is more practical than using radar and infrared detectors which are dependent extensively on computer processing.

This new technology can be utilized in developing sensors against collisions of vehicles on the road, monitoring and programming of video games. Especially, automotive industry can take the benefit of this collision avoidance system with its higher efficiency and lower cost. It is a visually stimulated motor control system (known as VSMC) which have two detectors for recognition of movement and a generator for motor commands. While the detectors evaluate the images, the generator changes them into motor commands.

In this research, the biology of locusts was scrutinized because of the fact that while seeing, the electrical and chemical signals enable Locusts a unique ability to realize and avoid forthcoming collisions.

More information about the research can be obtained from the essay of ‘Visually Stimulated Motor Control for a Robot with a Pair of LGMD Visual Neural Networks’ at the International Journal of Advanced Mechatronic Systems.