PENGARUH CONVOLUTIONAL NEURAL NETWORK UNTUK PROSES DETEKSI PENYAKIT PADA DAUN TOMAT

Abstract

Tomato plants are one of the plants that are often planted by farmers and are the main food requirement in
society. Tomato cultivation is often faced with disease problems that can attack the leaves, stems and fruit.
However, many farmers often face difficulties in overcoming this problem. To solve this problem, researchers
will use a web-based system that is able to classify images of tomato leaves. The system will process the image
first before training the CNN model. The resulting model will be used to classify images entered through the
website. Apart from that, this design also has several useful benefits. The results of the analysis of the model
show that there are challenges in distinguishing the characteristics of diseases in tomato plants, so that the
development of the CNN model experiences difficulties. Despite these difficulties, the CNN algorithm provides
an accuracy score of 0.9091. This number reflects the model's level of accuracy in classifying images into the
correct categories. From these results, it can be concluded that disease detection in tomato plants using the
CNN algorithm requires special effort and attention, especially in collecting representative datasets and
modeling optimal CNN architecture. A deeper understanding of the characteristics of diseases in tomato plants
also needs to be considered to increase the accuracy of model predictions. Although there is still room for
improvement, these results provide a basis for continuing to develop and improve disease detection models in
tomato plants using CNN approaches.