Mineral Grain Recognition
CNN-driven mineral grain recognition in water sand samples.
Tech Stack:
PythonTensorFlowCNN
Development Timeline
Project Kickoff
Data collection and model setup.
- Collected sand samples
- Preprocessed images
- CNN architecture designed
Data Preparation
Prepared dataset for training.
- Image augmentation
- Dataset split
- Labeling completed
Model Development
Model training and testing.
- Trained CNN model
- Achieved 85% accuracy
- Integrated predictions
Evaluation Phase
Evaluated and refined model.
- Hyperparameter tuning
- Error analysis
- Final validation