Application of Artificial Intelligence to Detect Lycorma delicatula

by Annalyse Kehs
Lycorma delicatula, Spotted Lanternfly, has been in Pennsylvania since 2014 and increasing their mass destruction to staple crops and habitats each year. Current methods to track population and eradication efforts are time-consuming and inaccurate. Machine learning offers a novel solution to these shortcomings. We present an object detection model trained to recognize nymphs on a sticky band and return the count of nymphs. Results from lab testing show +/- 5% accuracy compared to ground-truth counts.
Major: 
Biological Engineering
Exhibition Category: 
Health and Life Sciences
Exhibition Format: 
Poster Presentation
Campus: 
University Park
Faculty Sponsor: 
David Hughes
Location: 
Alumni Hall, HUB-Robeson Center
Poster Number: 
104