Mobile Application Development for Chili Disease Detection with Convolutional Neural Network

Mobile Application Development for Chili Disease Detection with Convolutional Neural Network

Chili Plant Disease, Convolutional Neural Network, System Mobile

The demand for chili continues to increase along with population growth and the industrial sector, but the supply is unstable due to weather factors such as high rainfall and humidity. This condition causes the spread of diseases in chili plants such as anthracnose fruit rot, begomovirus yellow virus, and leaf spots. This study aimsto develop a chili plant disease identification system and evaluate the accuracy of chili plant disease image classification. This study is expected to help experts provide recommendations for accurately controlling the spread of disease, assisting farmers in identifying diseases early,and improving the quality and quantity of chili plant harvests. The method used is Transfer Learning using the Convolutional Neural Network (CNN) MobileNet V2 architecture. The research stages include data collection, needs analysis, preprocessing design,model design, architecture stages, accuracy testing, system implementation, and system testing. These stages are carried out sequentially withoutany being skipped. Accuracy testing is calculated using a confusion matrix. In addition, the system that has been created will be tested by functional and user testing. Chili plant disease image data was obtained from chili plantations in Sumowono District, Semarang Regency. A dataset of 4,500 chili plant disease images was used, and 70% of the data was divided intotraining data and 30% for validation data. The accuracy results obtained were 99% in the training process and 94% in the validation process. Model evaluation using a new dataset of 150 chili plant disease images showed an accuracy result of 94%. Functional testing and user testing by 10 farmers produced an average value of 90. Thus, it can be concluded that the system can identify chili plant diseases well, as well as support agricultural activities and farmer needs

Sri Winiarti1, Itsnaini Irvina Khoirunnisa2, Norhudah Seman3

https://ijicom.respati.ac.id/index.php/ijicom/article/view/93/62

December 2024

Fajar bagus W

PDF

English

Text