Development of oil palm tree's quality estimation by using image processing techniques from an unmanned aerial vehicle

Abstract

     Oil palm is an important crop in Thailand, valuable particularly in industry. Oil palm is an important component in various manufacturing industries. As a result, farmers are producing increasing amounts of palm oil. The palm oil plant suitable is suitable for hot and humid climates and the product can be harvested throughout the year. It is a plant that generates a lot of income for farmers and the government has funds to support cultivation. However, monitoring the health of palm trees over large areas can be difficult for farmers. This research aims to create a monitoring system for the health of palm trees within the palm plantation and to help farmers to monitor palm trees more conveniently. The system used unmanned aerial vehicles to capture a large area of palm trees from above. The images were then used to analyze the health of palm trees with index VARI (Visual Atmospheric Resistance Index). The images were subjected to digital image processing technique and then the VARI index was calculated.The system could be distinguish four levels of palm quality used to inspect damaged palm trees within the palm plantation. The damage may have been caused by weather and nutrient problems, etc. The experimental results showed that the palm oil monitoring system had an accuracy about 91.18%.