Efficient Monitoring of Mountain Pine Beetle Outbreak Spots Using Artificial Intelligence Applied to Drone Thermal Imagery

Accurate and efficient detection of green-attack pines is critical for the success of Alberta’s MPB management program. This project will attempt to develop a novel method for the detection of green-attack trees by capturing thermal imagery and using artificial intelligence to distinguish newly-attacked pines from healthy ones. If successful, these results would next be operationalized to enable to early detection of green-attack trees that would advance the MPB management program.

Objective

To develop an algorithmic approach for differentiating between recently dead pine trees with green foliage and live healthy pine trees using artificial intelligence to analyze thermal images of tree crowns. We hypothesize that variation in heat dissipating from the foliage of green attack and healthy trees in the form of long-wave radiation can be detectable using thermal infrared sensors.