Early yield prediction in pear based on canopy LIDAR scanning

J. E. BENITEZ PICCINI; J. C. MAGDALENA; J.L. FIGUEROA; F.R. MASSON

Mar Del Plata

Congreso; 2017 XVII Workshop on Information Processing and Control (RPIC); 2017

The agricultural industry faces the challenge of increase yield and improve the food quality. This paper presents a study on the prediction of fruit production and also intends to detect other information that can be obtained based on the scans of fruit trees using a LIDAR sensor. We worked on a Williams pear orchard using a semi-automatic procedure in order to estimates tree volume. We compute the canopy area with a simple method based on the Gauss's formula at harvest and during tree dormancy. We correlates volume vs. weight of harvested fruits from each plant. Regression coefficients were obtained for individual and grouped data. At harvest time, we obtain a good relationship between the volume of the canopy and the production (r = 67). The information obtained is very valuable in terms of the general state of the crop, the evolution and distribution of production within the orchard, and to make adequate decisions about its management.