<![CDATA[Fingerprint identification is an important biometric method used in criminal investigations. However, at the Identification Unit of the Bali Regional Police, the minutiae extraction process is still done manually. This causes limitations in accuracy, processing speed, and dependence on the skills of the officers. This study develops an automatic minutiae detection system based on MinutiaeNet. The system is equipped with a graphical user interface and an SQLite database to store extraction results, using a Research and Development (R&D) approach. The system was tested on 55 fingerprint images, including plain and latent fingerprints. The results show that the system performs well, with an average Precision of 0.83, Recall of 0.88, and F1-Score of 0.85. The system works stably on plain fingerprints with clear ridge patterns but shows lower performance on latent fingerprints due to thin, incomplete ridges and noise, which increases False Negative results. Although over- detection occurs in some cases, the system still provides consistent results and significantly reduces processing time. The maximum processing time is about 1 minute and 3 seconds per image, which is much faster than manual minutiae extraction. This study shows that the proposed system can support forensic identification by improving efficiency and consistency in fingerprint analysis.]]>
Copyrights © 2026
