ABSTRACT Huanglongbing is present in tropical and subtropical regions and poses a significant challenge to global citrus production, as there is currently no cure for the disease. Timely detection is crucial for effectively managing affected trees and preventing disease propagation, thereby minimizing further losses for producers. This review presents an overview of the disease, illustrating the symptoms observed in affected trees and the economic impact of the disease in different regions of the world. One of the primary challenges is the ease with which it can be mistaken for other diseases by those lacking experience. To address this, techniques employing diverse machine learning algorithms have been utilized to discern and identify characteristics and patterns in orange trees' leaves, as this is one of the most conspicuous manifestations of the disease. This work summarizes several research works on HLB detection using Deep Learning algorithms and spectroscopy techniques in HLB images of leaves and describes the most used datasets. This represents a promising avenue for advancing the field. These techniques are integral to implementing smart agriculture, facilitating broader access to new technologies that could help prevent the spread of disease and reduce production losses. This review identifies a significant gap between laboratory‐scale accuracy and field‐level reliability, highlighting the urgent need for cross‐dataset validation and standardized environmental noise suppression.
Torres‐Galván et al. (Tue,) studied this question.