Detection of citrus diseases using artificial intelligence: A systematic review

Abstract

Early detection of citrus diseases is important for the global agricultural industry, facing threats such as Huanglongbing and canker. This study reviews the current status of the use of artificial intelligence to improve detection accuracy and speed. A systematic literature review was conducted from 2019 to 2023, using databases such as Scopus, IEEE Xplore and ACM, focusing on identifying the fruits studied, prevalent diseases, AI algorithms used and their accuracies, as well as technical challenges in implementing AI systems. The results highlight that oranges, lemons and mandarins are the most investigated fruits, with Huanglongbing, black spot and canker as the most studied diseases. AI algorithms such as Deep Neural Networks (DNN) and Adaboost show high accuracies, essential to improve disease detection. However, challenges include lack of labeled data, adaptation to different agricultural conditions, and effective integration in dynamic agricultural environments. This study reveals the need to advance data quality and algorithm adaptability to strengthen sustainability and efficiency in disease detection in citrus crops.