Recent advances in pest and disease recognition: a comprehensive review
Honglin Liu, Bisheng Zhan, Ruitong Fang, Yi Zhang, Yujiao Ma, Zexiang Shen, Qirong Mao
Abstract
Agricultural pests and diseases pose a severe threat to global food production, making timely and accurate recognition crucial for ensuring crop health and enhancing yields. With the rapid advancement and application of artificial intelligence (AI) across various scientific domains, its potential in pest and disease recognition remains only partially explored. Therefore, we conduct a comprehensive review, focusing on the latest progress in applying machine learning (ML), deep learning (DL), and multimodal technologies to pest and disease recognition in agriculture. It covers state-of-the-art techniques, benchmark datasets, and evaluation metrics relevant to this field. Additionally, the review offers an in-depth understanding of the strengths, challenges, and limitations of these methods. We also highlight several representative studies and conduct a comparative analysis of their performance. Finally, the paper provides detailed insights, proposes potential research directions, and concludes with reflections on future advancements.