Automatic Identification Fingerprint Based on Machine Learning Method

The fingerprint identification technology has been developed and applied effectively to security systems in financial transactions, personal information security, national security, and other fields. In this paper, we proposed the development of a fingerprint identification system based on image processing methods that clarify fingerprint contours, using machine learning methods to increase processing speed and increase the accuracy of the fingerprint identification process. The identification system consists of the following main steps: improving image quality and image segmentation to identify the fingerprint area, extracting features, and matching the database. The accuracy of the system reached 97.75% on the mixed high- and low-quality fingerprint database.