Application of K-Means Clustering for Detection Downy Mildew at Madura Corn Plant Using Digital Image Processing

Imron Rosyadi NR (Universitas Madura)
Erwin Prasetyowati (Universitas Madura)
Badar Said (Universitas Madura)
Syaiful Arifin (Universitas Madura)
Mohammad Syafiir Ridoni (Universitas Madura)

Tibuana : Journal of Applied Industrial Engineering
Vol 6 No 2 (2023): Tibuana

The development and cultivation of corn is necessary in line with the increasing consumption of food ingredients and industrial needs, especially food products made from corn. In the development of maize in Indonesia, the main obstacle is the disturbance of Plant Pest Organisms (OPT), especially diseases, one of which is downy mildew. This disease can be identified by a change in color, so we need a way to find out the difference between the color of healthy leaves and the color of leaves that have changed due to downy mildew. One solution that can be used is image processing. Therefore the aim of this study was to detect downy mildew based on leaf color in corn plants based on digital image processing, to produce precise and objective results. The algorithm used is the K-Means Clustering algorithm. This study uses 50 images of training data and 25 images of test data. Based on the simulation of downy mildew disease identification using K-Means Clustering it achieves an accuracy rate of 85%.