Klasifikasi Tingkat Kerapatan Tajuk Kelapa Sawit Menggunakan Metode Convolutional Neural Network (CNN) Berdasarkan Citra Drone.

Abstract

Monitoring the canopy density of oil palm plantations is essential for assessing plant health, optimizing plantation management, and improving productivity. Conventional monitoring methods generally rely on field observations, which are time-consuming, labor-intensive, and often less efficient for large-scale plantations. The advancement of unmanned aerial vehicle (UAV) technology and deep learning offers an alternative solution for automated canopy analysis. This study aims to classify oil palm canopy density levels using the Convolutional Neural Network (CNN) method based on drone imagery. The dataset was acquired using drone flights over oil palm plantation areas and subsequently preprocessed through cropping, resizing, and data augmentation techniques. The canopy density was categorized into three classes: low, medium, and high. A CNN model was developed and trained to automatically extract image features and perform classification. Model performance was evaluated using accuracy, precision, recall, and F1-score metrics. The experimental results indicate that the proposed CNN-based approach effectively classifies oil palm canopy density with high performance. The integration of drone imagery and deep learning provides a rapid, accurate, and efficient solution for supporting precision agriculture and sustainable plantation management.