An evaluation of a vision-based sensor performance in Huanglongbing disease identification

Abstract

Huanglongbing (HLB) disease is a critical infection which has dangerously affected the citrus production in Florida and has also been observed in California and Texas. No active treatment for this infection has been reported yet and the HLB infected tree will eventually die. However, early identification and removal of the HLB affected trees will secure the healthy trees in the grove. Accumulation of starch on infected leaves is an early symptom of HLB. Starch can rotate the polarisation plane of light in a certain waveband. A customised vision sensor was developed based on this characteristic to identify the HLB symptom. The vision sensor images were compared with the images captured by a colour camera to demonstrate the improvement achieved in this method. Also, the starch accumulation identification was studied for citrus leaves before and after being ground. The results showed an enhanced HLB identification performance using the developed vision sensor. • Accumulated starch can rotate the polarisation plane of light near 600 nm. • The customised vision sensor optimised the HLB infection identification. • Starch accumulation was more evident in ground citrus leaf. • The vision sensor could separate HLB infection from nutrient deficiency.