Uncovering radar signatures to detect landmines
Countries throughout the world have joined forces to try to defuse the effects of landmines, which kill and injure thousands of people each year.
Mizzou Engineering Professor Dominic Ho has contributed substantially to that effort during the nearly 10 years in which he has focused on landmine detection research, having already helped develop a ground-penetrating radar system that better detects landmines. Now Ho has begun working to enhance the radar system to minimize the number of false alarms it generates in the field, backed by a $116,500 U.S. Army grant he received in late May.
“We have different types of what we call clutter objects in the ground—such as metal debris or pieces of rock—that create false positives,” said Ho, an electrical and computer engineering professor. “We’re working on a way to determine which of the ‘positive’ signals really are landmines.”
Under test conditions, the radar system that Ho helped devise has reduced false alarms from a rate five years ago of 100 per 5,000 square meters to the current rate of less than four per 5,000 square meters. That’s roughly equivalent to reducing false landmine detections in a football field from 117 to four—an improvement that saves considerable digging time and labor, Ho said.
Yet the wide variety of clutter objects that often lies up to six inches underground diminishes the detection system’s efficiency. Ho plans to improve the real-world effectiveness of the detection system by further analyzing the signals sent by its radar, which is mounted on a vehicle’s front bumper.
That analysis, Ho believes, will reveal radar patterns that can be used to decide exactly what an underground object is. Landmine shapes, whether circular or square, will create a regular radar pattern that analysis can uncover, compared to the irregular radar patterns that rocks and other clutter create, he said.
Differences between landmines are among the challenges Ho faces in trying to establish landmine radar signatures. Not only do landmines vary in shape, but some are made of plastic, some are made of metal and some have protrusions.
“That makes for a lot of variations that I’m trying to figure out,” Ho said. “Moreover, the radar signatures must be detailed enough to rule out false positives, but flexible enough to correctly identify all types of landmines.”
Ho expects the project to run three years.