Determination of olive tree (Olea europaea L.) some dendrometric components from unmanned aerial vehicle (UAV) data with local extrema and multiresolution segmentation algorithms

Abstract

In this study, it was aimed to determine the dendrometric components of olive trees by using an unmanned aerial vehicle (UAV). The research was carried out in the olive groves of Akdeniz University Faculty of Agriculture. The study consists of the basic stages of acquisition, processing and analysis of UAV images. In the first stage, autonomous flight was performed with the UAV and digital images of the area were collected. In addition, at this stage, the number and height of olive trees in the area were determined by making local measurements. In the second stage, orthomosaic image, digital surface model (DSM) and digital terrain model (DTM) were produced by processing UAV images. At this stage, tree crown boundaries were determined by manual digitization over the orthomosaic image. Then, a canopy height model (CHM) was created to semi-automatically calculate the crown borders, number of trees and tree height values of olive trees. As a result of the evaluation of semi-automatic findings and ground measurements, the general accuracy in the determination of trees in the olive grove was 96.15%, the accuracy of the producer was 85.14% and the user accuracy was 81.82% in the determination of the tree crown boundaries. In addition, high correlations were obtained in the determination of tree crown area (r = 0.980) and tree height (r = 0.918). According to these results, it has been revealed that some dendrometric components of the olive tree can be determined quite successfully with the semi-automatically calculated data from the UAVs.