A vanadium dioxide metasurface lets the smart infrared sensor switch between edge-enhanced and full-detail views.
Researchers from the City University of New York, the University of Rochester, the University of Melbourne, and RMIT University have created a compact sensor using image-processing metasurfaces. This sensor can be equipped on drones to provide cost-effective, large-scale monitoring of crops for irrigation, fertilisation, and pest control purposes.
The research team has developed a sensor based on a vanadium dioxide filter system that can toggle between infrared edge detection and detailed imaging as needed. “Vanadium dioxide’s ability to change from an insulator to a metal with temperature shifts enables our devices to smartly adapt their imaging,” explains Madhu Bhaskaran, a co-author of the study. This transformation allows the device to switch from capturing just the outlines of an image to a full, unfiltered infrared image.
The concept behind the metasurface filter is to provide a flat-optic sensor that can be easily integrated into drones and satellites. This sensor offers advanced imaging capabilities typically requiring large, power-intensive digital systems and bulky lenses, all within a significantly smaller package. Such sensors are handy for extensive agricultural monitoring, signalling the need for adjustments in irrigation, fertilisation, or alerting to pest infestations.
According to the team, the filter system can be mass-produced using temperatures that align with current manufacturing techniques, potentially allowing for a quick transition from research to practical application, as co-author Shaban Sulejman suggests. Furthermore, RMIT University has a significant stake in the success of this innovation, holding a US patent and a pending Australian patent for their proprietary method of producing the vanadium dioxide film utilised in this project.
“While a few recent demonstrations have achieved analogue edge detection using metasurfaces, most of the devices demonstrated so far are static,” explains co-first and co-corresponding author Michele Cotrufo of the team’s creation. “Their functionality is fixed in time and cannot be dynamically altered or controlled. “Yet, the ability to dynamically reconfigure processing operations is key for metasurfaces to compete with digital image processing systems. This is what we have developed.”