Answers are grounded in this paper and cite their sources.
This approach detects fruit flies using a deep learning model, improving pest monitoring in agricultural environments.
Protecting crops from harmful insects is very important for increasing food production and supporting sustainable agriculture. In recent years, the use of smart devices and sensors has made it easier to collect large amounts of data from farms. However, this data must be processed, understood correctly to help farmers make better decisions. This study presents a deep learning-based method for detecting harmful fruit flies that affect fruit quality and crop production. A new dataset with four classes of fruit flies was created by combining several binary and multi-class datasets to improve training and testing. Four different detection models were trained, tested to find out which one performs best for detecting one, two, three, and four types of fruit flies. Several image pre-processing methods, such as YMC norm, homomorphic filtering, min-max normalization, and negative training were applied before model training. These techniques were used to improve the accuracy, reliability of the models. Among all the models, YOLOv5 showed the best results in terms of precision, recall, F1-score, and mean average precision (mAP). An ESP32-CAM-based electronic trap was also added to the system. It captured real-time images and sent instant alerts through WhatsApp and email, allowing quick action against pests. This system is low-cost, easy to use, and suitable for farmers in developing countries like Pakistan. In conclusion, this approach helps reduce the overuse of pesticides and supports the production of high-quality fruits. It also offers a strong base for creating smart and automatic pest monitoring systems for future farming needs.
No takes yet. Share an insight, caveat, or question.
Anwar et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: