Design and testing of an automated system for targeted spraying in orchards

Abstract

The research aims to demonstrate the basic system elements of a prototype automated orchard sprayer, which can deliver pesticide spray selectively with respect to the characteristics of the targets. The shape of the apple tree canopy was detected by a machine vision system using an RGB camera and appropriate image analysis. Information captured by RGB camera and processed by specific software was fed in real-time to a spraying arm, with three individually controlled sections, which adapt the pesticide spray flow to the canopy shape. The system allows variation in the liquid flow rate and volume of chemicals by means of controlled electric valves, whereby the amount of spray depends on the shape of the tree crowns. The current project focuses on developing the critical elements for spraying an individual tree and evaluates them in independent field experiments with the use of water sensitive papers (WSP). The demonstrated concept of precise application of pesticide sprays supports decreasing of the amount of delivered spray, thereby reducing both costs and environmental pollution.