microalgae-image

This repository contains the code and image material described in: Stegemüller et al. (2025) - Online monitoring of Haematococcus lacustris cell cycle using machine and deep learning techniques

The files are organized as follows:

• Build CNN contains the construction, optimization, training, validation and evaluation (SHAP) of the Convolutional Neural Network
• Use CNN can be used to clasiify images using a trained Convolutional Neural Network
• Build XGBoost contains the construction, optimization, training, validation and evaluation (SHAP) of the Gradient Boosting Decision tree model (XGBoost)
• Use XGBoost can be used to clasiify images using a trained XGBoost model

The Dataset contains the dataset used for model training and validation The file contains one subfolder for each class ('Greenflag', 'Greenround', 'Others', 'Redflag' and 'Redround')