Deep Learning Mask R-CNN and Template Matching Algorithm For Tree Counting Analysis of Oil Palm Trees (Case Study: East Tanjung Jabung District, Jambi Province)

Abstract Indonesia is one of the largest producers and exporters of palm oil in the world. According to the Central Bureau of Statistics (BPS) in 2022 palm oil production has increased to 46.82 million tons. Palm oil has great potential; thus, technology is needed to analyze its productivity. Therefore, calculating palm tree percentages is an important aspect of land monitoring, plant maintenance, and efficient production planning. However, manual palm tree counting is time-consuming and labor-intensive and tends to be less efficient in large plantations. In addition, the lack of accuracy in estimating the number of trees and the age of crops can affect land planning and management. Algorithm development in geomatics science can automatically calculate oil palm trees using aerial photo data combined with template matching and deep learning methods, which are expected to provide efficient and accurate solutions. In this study, 620 samples of oil palm trees were trained. From the visual interpretation results, 6212 trees were obtained; from the processing results using the deep learning method, 6359 trees were obtained, and using the template matching method, 6756 trees were obtained. Through the validation test using the confusion matrix, the overall accuracy of the deep learning method was 95.15%, and the overall accuracy obtained using the template matching method was 87.17%.