Abstract:
In recent years, lidar-based remote sensing has gained popularity in entomological studies due to its ability to non-invasively sense insects in their natural habitat. Ho...Show MoreMetadata
Abstract:
In recent years, lidar-based remote sensing has gained popularity in entomological studies due to its ability to non-invasively sense insects in their natural habitat. However, previous studies that combined entomological lidar and machine learning for insect classification tasks have all been performed under controlled laboratory conditions. In this study, we compared several machine learning algorithms' ability to detect insects in field data with a high class imbalance of 7667:1. Using a single-hidden-layer neural network, we detected 61.19% of the insects, and were able to discard 98.25% of the testing data. Compared to state-of-the-art field studies where researchers manually detect insects, our results are a significant step towards automated insect detection and classification in field experiments.
Published in: 2021 IEEE 31st International Workshop on Machine Learning for Signal Processing (MLSP)
Date of Conference: 25-28 October 2021
Date Added to IEEE Xplore: 15 November 2021
ISBN Information:
Print on Demand(PoD) ISSN: 1551-2541