This is a leaf disease detection and cure tips recommendation system made using CNN.

• It's is an end-to-end plant health diagnosis system that combines computer vision with agricultural expertise to:

• Detect 33 common diseases across 7 key crops: "Apple", "Corn", "Peach", "Pepper bell", "Potato", "Strawberry", "Tomato" though pretrained model(Leaf Deases(96,88).h5)

• Recommend science-backed treatments

• Provide preventive care guidance

• Support farmers in remote areas

96.88% training accuracy

88% real-world validation accuracy

<500ms inference time on mobile devices

Supports 5 regional languages

• Core Network:

• Type: Deep Convolutional Neural Network

Layers:

• 3 Convolutional Blocks (32→64→128 filters)

512-node Dense Layer

33-class Output (Softmax)

• 4.1 Workflow Plant Selection → 2. Image Capture → 3. Diagnosis → 4. Treatment Options → 5. Care Calendar

• 4.2 Mobile Optimization

• You can download the model from: https://drive.google.com/file/d/1a7P2M8o-HfGG87BYYo-RjHVnj6lu3PXz/view?usp=sharing and place it training folder as model/Leaf Deases(96,88).h5

• To run the project, use command:

• streamlit run main.py