Loading...
Cucumber Disease Identification Using Multiple Machine Learning Classifiers with a Pre-Trained VGG16 Model
Author(s)
Sutacha, Chalermkiat
Yodrot, Teerapon
Orachon, Teerapong
Orkweha, Khwanjit
Date Issued
January 1, 2024
Type
Conference Paper
Abstract
Agriculture is a foundational element of global food security and economic stability, yet it faces the perennial challenge of crop diseases, which can significantly impact yield and quality. Traditional disease detection methods, reliant on manual inspection, are labor-intensive and can lack precision. Addressing this, we propose a hybrid model that combines the strengths of Convolutional Neural Networks (CNNs) and machine learning classifiers for the automated detection of diseases in cucumber plants. Specifically, we employ the VGG16 network, renowned for its feature extraction capabilities, coupled with Support Vector Machine (SVM), Random Forest (RF), Decision Trees (DT), and K-Nearest Neighbors (KNN) classifiers. Our model is trained on a dataset consisting of high-resolution cucumber images, with 80% used for training and 20% for validation. Performance metrics such as accuracy, precision, recall, and F1-score are utilized for evaluation, with a particular emphasis on the F1-score for its balance between precision and recall. The SVM classifier emerges as the most effective, followed by RF, KNN, and DT. The results, analyzed through confusion matrices, validate the SVM's superior performance in accurately classifying the health status of cucumber plants. This study not only enhances disease detection accuracy but also suggests a direction for future agricultural practices, highlighting the potential integration of such models into real-time crop management systems.
Citation
Proceeding 12th International Electrical Engineering Congress Smart Factory and Intelligent Technology for Tomorrow Ieecon 2024, 2024