Juhanas, Ade (2026) Identifikasi Penyakit Tanaman Di Buah Matoa Menggunakan Metode Convolutional Neural Network (CNN). Other thesis, Universitas Islam Riau.

Text skripsi_203510132_watermark.pdf - Published Version Restricted to Registered users only Download (3MB) | Request a copy

Abstract

This study aims to identify diseases in matoa (Pometia pinnata) plant leaves using an image-based Convolutional Neural Network (CNN) method. The dataset used consists of matoa leaf images classified into three classes: healthy leaves, powdery mildew, and leaf rust (rust). The data were obtained from field image acquisition and relevant supporting sources, followed by preprocessing stages including resizing, normalization, and data splitting into training and testing sets with an 80:20 ratio. The CNN model was trained using the VGG16 architecture with a transfer learning approach and several training parameter scenarios employing Adam and AdamW optimizers. Model performance was evaluated using accuracy, precision, recall, and F1-score metrics. The results show that the CNN model with the VGG16 architecture achieved the best performance with the highest accuracy of 95.75% on the testing data. The best-performing model was then implemented into a web-based application using the Flask framework to assist in the rapid and accurate identification of matoa plant leaf diseases.

Actions (login required)

View Item