Jasper Siebring (the Netherlands)
There is a growing demand in both food quality and quantity, but as of now, one-third of all food produced for human consumption is lost with pests and other pathogens accounting for roughly 40% of pre-harvest loss in potatoes. Pathogens in potato plants, like the Erwinia bacteria and the PVYNTN virus for example, exhibit symptoms of varying severity that are not easily captured by pixel-based classes (as these ignore shape, texture, and context in general). The aim of this research is to develop an object-based image analysis (OBIA) method for trait retrieval of individual potato plants that maximizes information output from UAV-based RGB VHR imagery and its derivatives, to be used for disease detection of the Solanum tuberosum. The approach proposed can be decomposed in two steps: object-based approximation of potato plants using an optimized implementation of Large Scale Mean-Shift Segmentation (LSMSS), and classification of disease within said approximations using a set morphological features computed from their associative objects. The proposed approach was proven to be viable as the associative Random Forest model detected presence of the Erwinia bacteria and potato virus Y with a maximum F1 score of 0.75 and an average MCC score of 0.47.