Development of Aerial Online Intelligent Plant Monitoring System for Oil Palm (Elaeis guineensis Jacq.) Performance to External Stimuli

Abstract

Researches in bio-robotics fields have been done en-masse. Development in intelligence monitoring systems for agricultural application have unfold the possibility to observe individual plant response upon receiving external stimuli. In this study, artificial bio pores, 30 cm in diameter, were introduced to the ranges of oil palm trees in three commercial plantations. Various applications methods of bio pores, in particular depths and numbers per plant were investigated. The bio pores drilled around the root zone of the trees using an earth auger, and filled with chopped semi-decomposed fronds and midribs from the plantation maintenance (pruning). A robotic quadcopter drone with 2.7K camera, operated with pre-set flight-plan, employed to record the crown image of oil palm trees under observation. The drone flown at the altitude of 23±0.1 meters above the crown, recording each crown individually. Focus and setting of drone’s cameras was set to automatic, enabling unbiased image recording. The weather conditions (sun radiation, cloud covering, wing speed) upon images recording were measured and recorded. When recording the images, the drone’s GPS-assisted hovering system maintained its position in both axes (horizontal and vertical), producing identical image acquisition for each crown. All plants’ crown was observed at 0, 30, 60, and 120 days after bio pores introduced. Image processing software was developed to segment and extract vegetation index (Vis) information from the images. Plants’ morphological conditions (height, radial, and new leaf) were measured and analyse by statistical methods to understand various bio pores applications influences to plants development. Crown images were processed, and its features extracted and correlated with chlorophyll in leaves. Models developed to predict chlorophyll contents (A, B, and Total) in crown and Vis analyses methods were used to compare individual plant responding to this external stimulus by means of rotational-pivot charts. Results showed that intensive bio pores introduction promote plant’s radial development and the emergence of new leaves. Furthermore, chlorophylls contents in leaves of plants with substantial bio pores applications were greater compared to normal plants.  Models showed that optical features extracted from crown images obtained high coefficient of correlation (R2) with leaves chlorophyll contents. This study has paved the way for wisdom agricultural application in Indonesian oil palm industry.