Real-Time Tree-Foliage Surface Estimation Using a Ground Laser Scanner

The optimization of most pesticide and fertilizer applications is based on overall grove conditions. In this paper, we propose a measurement system to estimate the foliage surface of a tree crop. The system is based on a ground laser scanner that estimates the volume of the trees and then extrapolates their leaf area using simple and fast algorithms to allow true real-time operation. Tests with pear trees demonstrated that the relation between the volume and the foliage can be interpreted as linear with a coefficient of correlation $(R)$ of 0.81, and the foliage surface can be estimated from this volume with an average error less than 6%.

I. Introduction

Traditional pesticide and fertilizer application in the treatment of groves is based on uniform application, which causes significant material loss (Fig. 1) [1]. The measurement of the overall grove conditions can be used to optimize the application, to provide economic benefits for growers, and to reduce pollution in the environment [2]. The measurement of tree-canopy structural characteristics, such as volume and foliage, can be carried out by means of several detection principles, such as image-analysis techniques, stereoscopy photography, analysis of the light spectrum, the infrared thermograph, ultrasonic sensors, and light detection and ranging (LIDAR) laser sensors. Fig. 1.

Pesticide application.