DeepPestNet: High-Accuracy Deep Learning for Crop Pest Detection and Classification in Precision Agriculture
作者
M. Harshini,A. Pavani,T. Katyayini,P. Krishna Reddy,M.Varshith Kumar
出处
期刊: [Institute of Electrical and Electronics Engineers] 日期:2026-01-27卷期号:: 1297-1302
标识
DOI:10.1109/icipcn67432.2026.11438986
摘要
Insect infestations pose a great impact on agricultural production and are many times ignored, particularly in terms of human resource availability and the efficiency of manual inspection techniques. To address this issue, this paper presents DeepPestNet, a light and efficient deep learning framework, which accurately identifies pests in real-world agricultural field environments. In this paper, a new approach is proposed, which combines convolutional feature extraction, feature pyramid networks, and attention mechanisms to effectively address real-world challenges like changes in illumination, background noise, and small and overlapped pests. DeepPestNet will be trained and validated on a large-scale AgriPest dataset of approximately 50,000 images, with over 260,000 labeled pest bounding boxes, and a tremendous amount of data augmentation techniques. The experimental investigation of the proposed framework demonstrates the outperformance of state-of-the-art deep learning techniques in terms of accuracy, efficiency, and generalization abilities, while still remaining computationally efficient to be deployed on resource-limited mobile and edge computing environments. This paper addresses sustainable and efficient agricultural pest management techniques, as it assists in decision-making and reduces pesticide use.