Paper B14

Unmanned Aerial Vehicle Application for Plantation Mapping
and Automatic Oil Palm Trees Counting
on Oil Palm Plantation Management

Hermantoro*,1, Andiko Putro Suryotomo2, Arief Ika Uktoro1, and
Rengga Arnalis Renjani3
1Department of Agricultural Engineering, Stiper Agricultural University
2Department of Information Technology, University of Technology Yogyakarta
3Department of Mechanical and Biosystem Engineering , Bogor Agricultural University
Email: her_mantr@yahoo.com*

Abstract

Indonesian oil palm plantations are reaching a total area of about 10.9 million ha in 2014, with production reaching 31.4 million tons. To maintain the productivity and sustainability of oil palm plantations, effective and efficient management are needed. Supported by the rapid developments information and communication technology, UAV (drone), computer vision system, and spatial modeling is the right choice to support the management of oil palm plantations. The counting of oil palm trees are an important part in the management of oil palm plantations, with regard to fertilization, harvesting, and monitoring tree conditions. This paper discusses the application of drones for plantation aerial image data acquisition and analysis, followed by calculation of the number of oil palm trees automatically by using software based on Viola-Jones algorithm. The image of oil palm plantations is obtained from the data processing of aerial photographs using the drone and then through mosaic image and composite processing. The resulting aerial image of the oil palm plantations then will be used to develop the automatic oil palm trees counting software. Based on accuracy testing, the software achieved up to 96% accuracy.

Keywords: Drone, oil palm, tree counting, Viola-Jones.

Full Text PDF of B14 Hermantoro et al