Identification of soybean planting gaps using machine learning

Abstract

The identification of planting gaps is essential for optimizing crop management in precision agriculture. Traditional methods, such as manual scouting, are limited in scale and precision. This study evaluates the performance of three machine learning algorithms—Decision Trees, Support Vector Machines (SVM), and Multilayer Perceptron (MLP) Neural Networks—for classifying planting gaps in soybean fields using UAV imagery during the V4 growth stage. The Neural Network and SVM models demonstrated similar results, with the Neural Network achieving an AUC of 0.984, accuracy of 94.5 %, F1 score of 0.945, precision of 94.5 %, and recall of 94.5 %. The SVM model with a Polynomial kernel achieved an AUC of 0.989, accuracy of 95.5 %, F1 score of 0.955, precision of 95.5 %, and recall of 95.5 %. In contrast, the Decision Tree model performed lower, with an AUC of 0.805 and accuracy of 79 %. These results demonstrate the effectiveness of machine learning algorithms, particularly Neural Networks and SVM, in improving planting gap detection, contributing to more precise crop management decisions.