Identification of Citrus Huanglongbing Disease at the Pre-Symptomatic Stage Using Polarized Imaging Technique

摘要

Huanglongbing (HLB), also known as yellow shoot or citrus greening, is a devastating disease, associated with 'Candidatus Liberibacter asiaticus (CLas).' Unfortunately, there is no effective cure for this disease, and since a tree is infected, it will eventually die. HLB has negatively affected citrus industries in many countries in Africa, Asia, and South America. In 2005, HLB was confirmed in a groove in South Florida, and after a few years, the entire state of Florida got infected. Timely detection and removal of HLB-affected trees are substantially necessary for managing the disease. Accumulation of starch that appears as yellow/light-green islands on the citrus leaf is known as an early diagnostic symptom of HLB. A polarized imaging system was used to follow the hyper-accumulation of starch in leaves of graft-inoculated citrus plants. Time-lapse images of citrus leaves were acquired in a weekly manner after plants were grafted with a disk from another HLB-affected leaf to follow the variation in the images' gray values at different locations on a leaf. An image analysis algorithm was developed to highlight the regions of citrus leaf samples with an increase in average gray value as an indication of starch accumulation in the pre-symptomatic stage. The proposed method was able to detect successfully any local increase in the image gray values which associates with hyper-accumulation of starch, and consequently, the existence of CLas.