A master’s thesis at the College of Computer Science and Information Technology at  University of Basrah discussed techniques for classifying digital fingerprints. The thesis of researcher Jinan Abdul Karim Abdul Hassan presents a system that reduces the computational cost of searching for a fingerprint image in large data sets by dividing it into categories to reduce the search set and the number of comparisons and provide  Time to research and assist in developing a fingerprint recognition system. This system improves the rudder as it uses deep learning with pre-trained CNN models to extract and classify features. And improve performance by using pre-trained DCNN models for (feature extraction) and machine learning algorithms for dimensionality reduction (feature reduction), then (classification) algorithms, to accelerate and improve fingerprint recognition accuracy and avoid overfitting by reducing parameters.