New AI combo spots apple leaf disease with near-perfect accuracy
Gautam V, Kaur G, Rani J, Rajendran S, Ghantasala GSP
Crop Improvement
If you've got an apple tree in your backyard with mysterious spotted leaves, this kind of tool could eventually live in a phone app that tells you exactly what's wrong and where, before the problem spreads to the whole tree.
Scientists stitched together several AI techniques, one that traces the exact outline of a diseased leaf spot, another that cleans up background clutter, and a smart feature-matching system, then fed it all into a fast detection model. Tested on thousands of real apple leaf photos across five different disease types, the combined system correctly identified the disease nearly every single time, and it even showed its work by highlighting the actual infected tissue it used to make the call.
Key Findings
Achieved 99.95% classification accuracy across five apple leaf disease classes
Combined U-Net segmentation, EfficientNet, and an attention-based autoencoder with Canonical Correlation Analysis to fuse features before YOLO-based detection
Validated with five-fold cross-validation and paired t-tests (p < 0.05) across multiple public apple leaf datasets (FGVC7, PlantVillage, Kaggle, ATLDSD)
chevron_right Technical Summary
Researchers built an AI system that automatically finds and classifies diseased spots on apple leaves with over 99% accuracy, even in messy real-world photos with bad lighting or cluttered backgrounds. This could help farmers catch and treat crop disease much earlier, saving both yield and money.
Abstract Preview
Original paper
Comprehensive plant disease classification and severity estimation for sustainable farming via automatic segmentation and multi-scale feature fusion.
Detecting plant leaf diseases at an early stage is one of the most important requirements for sustainable agriculture, increasing crop productivity, and achieving the global Sustainable Development...
open_in_new Read full abstractAbstract copyright held by the original publisher.
Species Mentioned
Was this useful?
Want to tell us more? (optional)
Thanks for the note!
Something went wrong — please try again.
Too many submissions. Try again in an hour.
Gene editing removes 97% of celiac-triggering proteins from bread wheat
It could mean that people with celiac disease — roughly 1 in 100 worldwide — may one day safely eat bread made from real wheat, without sacrificing the taste...
An apple is the round, edible fruit of an apple tree. Fruit trees of the orchard or domestic apple, the most widely grown in the genus, are cultivated worldwide. The tree originated in Central Asia, where its wild ancestor, Malus sieversii, is still found. Apples have been grown for thousands of ...