This repository contains the Python-based automated image processing pipeline developed to predict mineral liberation behavior from polished section SEM images of chalcopyrite and pyrite.
The project implements an automated workflow using Python to preprocess SEM images, segment mineral grains, and quantitatively analyze grain size distributions to estimate liberation efficiency. This approach supports early-stage mineral processing design by providing predictive insights from polished section textures without requiring fully liberated particle samples.
- Image preprocessing (grayscale conversion, Gaussian filtering)
- Adaptive thresholding and morphological operations
- Connected component analysis for grain detection
- Grain size distribution and liberation percentage calculation
- Batch processing support
- Install required Python packages: pip install opencv-python numpy matplotlib pandas seaborn jupyter
- Open the Jupyter notebook: jupyter notebook MineralLiberationProject/liberation_sem
data/: SEM images and input data filesoutputs/: Results such as CSV files and labeled imagesrequirements.txt: List of Python packages neededLICENSE: License information