Perbandingan Algoritma Klasifikasi untuk Mendeteksi Kebutuhan Nitrogen Tanaman Padi Berdasarkan Data Citra Multi-spectral Drone. (Indonesian)

Comparison of Classification Algorithms for Detecting Nitrogen Needs in Rice Plants Based on Multi-spectral Drone Images. (English)

Optimizing the use of nitrogen (N) fertilizer is very important for increasing the productivity of rice plants. Traditionally, determining the appropriate fertilizer amount involves employing a leaf color chart (BWD), where farmers manually match rice leaf colors to a color scale. However, this method proves time-consuming. A promising strategy to boost efficiency involves utilizing a Multi-spectral Drone to capture multispectral images for precise N fertilizer assessment. This study aims to compare various classification algorithms for modeling N fertilizer needs from multispectral image data, using ground truth from BWD scaling. The classification algorithms tested include decision tree (DT), artificial neural network (ANN), support vector machine (SVM), random forest (RF), and k-nearest neighbor (KNN). The performance of the five classification algorithms is evaluated based on the metrics of accuracy, recall, precision, and F1 score. The results show that the classification model that provides the best performance is the decision tree (DT) algorithm, both in treatment without normalization and balancing, as well as in treatment involving normalization and balancing with accuracy, recall, precision, and F1-score values above 90%. [ABSTRACT FROM AUTHOR]

Optimalisasi penggunaan pupuk nitrogen (N) menjadi hal yang krusial dalam upaya meningkatkan produktivitas tanaman padi. Penentuan jumlah pupuk yang tepat untuk tanaman padi sering kali dilakukan dengan bagan bagan warna daun (BWD), di mana petani secara manual mencocokkan warna daun padi dengan skala warna pada BWD namun, proses ini terbukti memakan waktu. Salah satu strategi untuk meningkatkan efisiensi penentuan kebutuhan pupuk N adalah dengan menggunakan multi-spectral drone. Drone digunakan untuk mengambil citra multispectral, yang selanjutnya digunakan untuk menentukan kebutuhan pupuk N. Penelitian ini bertujuan membandingkan beberapa algoritma klasifikasi untuk memodelkan kebutuhan pupuk N dari data citra multispectral, dengan menggunakan ground truth dari penskalaan BWD. Algoritma klasifikasi yang diuji antara lain decision tree (DT), artificial neural network (ANN), support vector machine (SVM), random forest (RF), dan k-nearest neighbour (KNN). Kinerja kelima algoritma klasifikasi dievaluasi berdasarkan metrik accuracy, recall, precision dan F1 score. Hasil penelitian menunjukkan bahwa model klasifikasi yang memberikan kinerja terbaik adalah algoritma decision tree (DT) baik dalam perlakuan tanpa normalisasi dan balancing, maupun normalisasi dan balancing dengan nilai accuracy, recall, precision, dan F1-score di atas 90%. [ABSTRACT FROM AUTHOR]

Copyright of Jurnal Ilmu Komputer dan Agri-Informatika is the property of IPB University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)