Upload a leaf image
Choose a durian leaf photo in the dedicated analysis workspace.
AI-assisted durian leaf intelligence
See the signal in every leaf.
Analyse a durian leaf image with YOLO localisation and ConvNeXt-Tiny classification, then review the predicted condition, confidence, class probabilities, and available treatment guidance.
Decision support for four currently supported classes—not a guaranteed professional diagnosis.
How DurianCare works
From image to a focused result
A four-stage pipeline keeps the analysis understandable from upload through guidance.
Localise the leaf region
YOLO detects and crops the leaf area to focus the next stage.
Classify the condition
ConvNeXt-Tiny compares the crop across the four supported output classes.
Review the report
See confidence, probabilities, evidence, and available guidance.
Useful information, without overclaiming
DurianCare makes model output easier to review while keeping uncertainty visible.
Focused analysis
Classification concentrates on the detected leaf region rather than the full scene.
Transparent confidence
Confidence and all four class probabilities are presented alongside the result.
Available guidance
Existing treatment guidance or a Healthy monitoring message appears where available.
History and safeguards
Signed-in users can revisit records, while unclear and invalid submissions stay visibly flagged.
Four current model classes
When an image can be analysed, DurianCare reports one of these exact supported outputs.
Algal
Supported disease class
Blight
Supported disease class
Healthy
Supported healthy class
Phomopsis
Supported disease class
Built for people learning from leaf health data
- Farmers
- Agriculture students
- Researchers
- General users interested in plant disease detection
Ready to review a leaf?
Use the analysis workspace to upload an image and receive an AI-assisted report.