Crop Disease Detection Using Image Processing
Anushree Wagde , Pritee Raut , Vansh Karande , Rohit Yeskar , Chetan Talwekar, Durgesh Siriya
DOI : 10.46335/IJIES.2026.11.2.2
Abstract – Farming is super important for food and money in our country, but when crops get sick, farmers lose a lot — sometimes even 30–40% of their yield. Checking leaves manually takes too much time, needs experience, and most farmers can’t do it perfectly. So in this project I made a system that can automatically find diseases in crop leaves just by looking at their photos. It uses normal image processing steps: cleaning the image, finding the bad parts, pulling out important features like color changes, texture, and shape, and then using machine learning to tell what disease it is. I added some smart tricks like adaptive thresholding and combined two classifiers (SVM + Random Forest) so it works better. On my test with around 5000 leaf photos, it got 92% accuracy — pretty good I think! This kind of system can really help farmers catch diseases early using just their phone camera.
