AERIAL THERMAL INFRARED DETECTION OF PFM-1 PLASTIC BUTTERFLY MINES
We conducted a series of field detection experiments meant to simulate the mountainous terrains where PFM-1 mines were historically deployed and remain in place. In our tests, 18 inert PFM-1 mines along with the aluminum KSF-1 casing were randomly dispersed to mimic an ellipsoidal minefield of 8-10 x 18-20 m dimensions in a de-vegetated rubble yard at Chenango Valley State Park (New York State). We collected multiple thermal infrared imagery datasets focused on these model minefields with the FLIR Vue Pro R attached to the 3DR Solo UAV flying at approximately at 10 m. We identified different environmental variables to constrain the optimal time of day and daily temperature variations to reveal presence of these plastic UXOs. We show that in the early-morning hours when thermal inertia is greatest, the PFM-1 mines can be detected based on their differential thermal inertia. After processing datasets, we found that an increase in moisture content, finer grain sediment and surface laid mines are optimal conditions to detect the PFM-1. We anticipate that following further development, this remote sensing method can aid in significantly reducing the cost and time associated with landmine remediation in post-conflict nations.