Precision Spraying Model Based on Kinect Sensor for Orchard Applications

Abstract

ABSTRACT. This study describes the use of an integrated Kinect video sensor in a spray system to capture color and depth information to improve the efficiency of spraying in an orchard. Equations for calculating the optimal spray distance and pattern based on data collected by the video sensor are proposed, in conjunction with the leaf wall area density, to address the difficulty in estimating the dose of sprayed chemicals. First, to estimate the optimal distance and route, this study designs a spray system using the Kinect integrated video sensor and proposes an equation for calculating the leaf wall distance of fruit trees for spraying. Second, to control the dose of sprayed chemicals by adjusting the nozzle angle and rate of flow, this study proposes the concept of leaf wall area density and analyzes the leaf area shapes of peach, apricot and grapevine, which are fan shaped, round and square, respectively. Finally, the results of the experiment on peach trees, apricot trees and grapevines demonstrated that the intelligent orchard chemical precision spray model established based on the spray distance and the leaf wall area density can adjust the system to the optimal spray distance and dose, which can improve efficiency in spraying chemicals, increasing the economic benefits for fruit farmers. Keywords: Integrated sensor, Kinect, Leaf wall area density, Optimal spray distance.