Detection of palm oil health through multispectral UAV platform

• Main Authors: Rafezall, Carl Marlyana, Darwin, Norhadija, Mohd. Ariff, Mohd. Farid, Majid, Zulkepli
• Format: Conference or Workshop Item
• Published: 2020
• Subjects: SB Plant culture
• Online Access: http://eprints.utm.my/id/eprint/92269/; http://dx.doi.org/10.1109/ICSET51301.2020.9265374
• Institution: Universiti Teknologi Malaysia

In 2018, 19.5 million tonnes of crude palm oil from a planted area of 5.8 million have been produced in Malaysia. RM67.5 billion has been earned from palm oil export which contributes to one of Malaysia's gross domestic products (GDP). Unfortunately, palm oil cultivation is suffering from the destructive disease known as Ganoderma Boninense which causes the basal stem rot (BSR) disease. The objectives of this study are to provide a low-cost monitoring method for palm oil mapping and to analyse the health condition of palm oil trees through the spectral reflectance curve. The implementation of Unmanned Aerial Vehicles (UAV) that attach with the multispectral camera was used as the platform in data acquisition to monitor the condition of palm oil health. The multispectral UAV can determine the plant health to identify the early stage of the disease by extracting the palm oil leaf spectral reflectance curve. The range of Normalized Difference Vegetation Index (NDVI) of 0 to 1 is for good condition, while 0 to -1 is for dead or extremely stressed condition of palm oil. The result of spectral reflectance curve from multispectral camera was compared and analysed with the reading from spectroradiometer observation. Besides, flight planning and camera calibration were done before the data acquisition process. Ground control points (GCP) were established using GNSS method at the survey area for the geo-referencing process. The area of the study is located at the palm oil plantation of Sime Darby Plantation, Kulai. The outcome of this study is to produce a Normalized Difference Vegetation Index (NDVI) map as the final output.