PoS - Proceedings of Science
Volume 458 - International Symposium on Grids & Clouds (ISGC) 2024 (ISGC2024) - Artificial Intelligence (AI)
Deep learning approaches for prevention of Japanese local monkey trespassing in a sweet potato field
A. Sang-ngenchai*, H. Ogawa and M. Nakazawa
Full text: pdf
Published on: October 29, 2024
Abstract
In rural areas of Hakusan, Ishikawa Prefecture, Japan, the local monkey population has been causing damage to sweet potato farms and surrounding lands. To address this issue, a prototype system has been developed to assist farmers in protecting their fields during the months of September through November in both 2022 and 2023. The system is based on deep-learning models utilizing the you-only-look-once (YOLO) algorithm to classify and localize images of the local monkeys simultaneously. Real-time detection of live monkeys is possible using a Streaming Protocol Camera (RTSP), with the system automatically notifying farmer group members via the Line application when a monkey is detected. The system was trained using data collected from trap cameras placed around the sweet potato fields and successfully operated on a Windows PC with 32 GB RAM, a 64-bit Operating System, and an Intel(R) Core(TM) i9-9900K processor with an Nvidia GeForce RTX 3080 10 GB graphics processing unit (GPU). Performance metrics based on k-fold cross-validation showed precision, recall, and AP@0.5 values of 0.7310, 0.8462, and 0.7421, respectively, indicating high accuracy and classification performance using a real-time streaming camera. By alerting farmers in advance, the system can prevent damage caused by monkeys in sweet potato fields.
DOI: https://doi.org/10.22323/1.458.0033
How to cite

Metadata are provided both in "article" format (very similar to INSPIRE) as this helps creating very compact bibliographies which can be beneficial to authors and readers, and in "proceeding" format which is more detailed and complete.

Open Access
Creative Commons LicenseCopyright owned by the author(s) under the term of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.