A comprehensive review and bibliometric analysis of deep learning applications in plant disease detection

Abstract Agriculture is a major component of many countries’ economies worldwide, especially in developing nations. The major challenge farmers face is the rapid spread of plant diseases, which often goes unnoticed and causes major yield and economic losses. Traditional methods such as human inspection and laboratory testing are often time-consuming, inaccurate, labor-intensive, and unsuitable for large-scale farming. To overcome these challenges, advanced technologies like Machine learning (ML) and Deep Learning (DL) have emerged in recent times, providing early and accurate detection of plant diseases. This paper focuses on recent advances in applying deep learning techniques, such as Convolutional Neural Networks (CNNs), to detect plant diseases. Across 1081 publications from 2014 to 2024, studies show that CNNs are well suited to extracting and analyzing features from plant images and clearly distinguishing healthy from diseased specimens. The models discussed in the surveyed literature generally report performance in terms of key parameters such as accuracy, precision, recall, and F1-score, which this review compiles and discusses rather than benchmarking through new experiments. The results show high accuracy and strong potential for real-world agricultural applications. The bibliometric analysis reveals growing interest and collaborative efforts in the research domain, especially among institutions specializing in artificial intelligence and agriculture. Moreover, the results support the practical application of deep learning in agriculture for plant health monitoring and highlight the importance of comprehensive studies in solving agricultural problems. The study provides a foundation for further innovations in automated plant disease detection and supports the transition to smart farming solutions.