Detecting Disease in Plant using ANN and CNN models in Python
Apurba Adhikari
Data Scientist
ML Engineer
Python
Project Aim:
Developed a solution to help farmers detect plant infections using image capture technology focused on plant leaves.
Workflow:
1. Data Collection: Gathered 4000+ images of plant leaves from three varieties.
2. Soil Sensor Data: Recorded soil sensor data to supplement image analysis.
3. Model Development: Created ML and DL models to detect plant diseases using image and sensor data.
4. Dashboard Creation: Built a visualization dashboard to present analysis and actionable insights for farmers.
Result:
Dataset: Included Banana and Cotton plants, each with two diseases.
ANN Model: Trained with soil sensor data for 150 epochs, achieving 92.98% accuracy.
CNN Models:
Banana: Trained for 97 epochs, achieving 92.99% accuracy.
Cotton: Trained for 100 epochs, achieving 95.96% accuracy.
