Fast Detection of Olive Trees Affected by Xylella Fastidiosa from UAVs Using Multispectral Imaging.

Area of Science:

• Plant Pathology
• Remote Sensing
• Agricultural Engineering

Background:

• Xylella fastidiosa (Xf) is a significant bacterial pathogen impacting over 350 plant species, including olive trees.
• Olive quick decline syndrome (OQDS), caused by Xf, poses a severe threat to olive cultivation, particularly in Southern Italy.
• Effective management strategies rely on early detection of Xf and control of its insect vectors.

Purpose of the Study:

• To develop and evaluate a fast detection method for Xylella fastidiosa symptoms in olive trees using UAV-based image processing.
• To assess the efficacy of a novel segmentation algorithm and linear discriminant analysis for identifying OQDS-affected trees.

Main Methods:

• Acquisition of high-resolution visible and multispectral images using a multirotor unmanned aerial vehicle (UAV).
• Application of a new image segmentation algorithm to isolate individual trees from acquired imagery.
• Classification of segmented trees using linear discriminant analysis to detect Xylella fastidiosa infection.

Main Results:

• The segmentation algorithm achieved a mean Sørensen-Dice similarity coefficient of approximately 70%, indicating successful tree isolation.
• The classification model demonstrated high performance with 98% sensitivity and 93% precision in detecting OQDS-affected trees.
• These results suggest a low rate of false positives and false negatives in identifying infected olive trees.

Conclusions:

• UAV-based image processing, combined with advanced segmentation and classification techniques, provides an effective tool for the rapid detection of Xylella fastidiosa.
• This approach holds significant potential for early and accurate identification of olive quick decline syndrome, aiding in disease management efforts.