- Literature survey of typical images quality features.
- Selection of broad range of features to describe the image quality.
- Training of a machine learning algorithm based on the selected features for distinguishing different image quality levels.
- Evaluation of trained algorithm.
- Critical review and summary of the achieved results