TJAHJADI, BERNARD FEBRIYANTO (2025) PLANT DISEASE DETECTION IN INDONESIAN STAPLE CROPS. S1 thesis, UNIVERSITAS KATOLIK SOEGIJAPRANATA.

Preview	Text 18.K1.0059_BERNARD FEBRIYANTO.T_COVER.pdf Download (472kB) | Preview
	Text 18.K1.0059_BERNARD FEBRIYANTO.T_BAB I.pdf Restricted to Registered users only Download (480kB)
	Text 18.K1.0059_BERNARD FEBRIYANTO.T_BAB II.pdf Restricted to Registered users only Download (481kB)
	Text 18.K1.0059_BERNARD FEBRIYANTO.T_BAB III.pdf Restricted to Registered users only Download (728kB)
	Text 18.K1.0059_BERNARD FEBRIYANTO.T_BAB IV.pdf Restricted to Registered users only Download (1MB)
	Text 18.K1.0059_BERNARD FEBRIYANTO.T_BAB V.pdf Restricted to Registered users only Download (502kB)
Preview	Text 18.K1.0059_BERNARD FEBRIYANTO.T_DAFPUS.pdf Download (409kB) | Preview
	Text 18.K1.0059_BERNARD FEBRIYANTO.T_LAMP.pdf Restricted to Registered users only Download (464kB)

Abstract

Plant diseases are one of the main challenges in the agricultural sector, especially in staple crops such as rice, corn, soybeans and cassava. This disease can cause significant economic losses if not detected early. This research aims to develop a plant disease detection system using Convolutional Neural Network (CNN). The dataset used includes 15,000 images of plants in healthy and diseased conditions, obtained from various online sources. The CNN model was designed with three convolution layers, ReLU activation function, and Nadam optimization, resulting in a validation accuracy of 85.02%. The system is implemented in a simple user interface-based application, allowing users to upload plant images and receive diagnostic results instantly, including relevant solution recommendations. The research results show that CNN-based technology can be an effective tool for detecting plant diseases with high accuracy. This research makes a significant contribution in supporting precision agricultural innovation. Suggestions for future research include collecting richer local datasets, optimizing model architecture, and integrating with IoT technology for real-time monitoring.

Actions (login required)

View Item