Databases

Fingerprint databases are curated collections of fingerprint images used for developing, testing, and benchmarking fingerprint recognition systems. They are essential for advancing research and ensuring reliable fingerprint identification

  • IAB Fingerprint Databeses:
    https://iab-rubric.org/index.php/mslfd

    The IAB-RUBRIC fingerprint dataset portal assembles a broad suite of modern research collections, capturing real-world challenges such as cross-device variation, presentation attacks and low-quality forensic impressions:
  • IIITD Multi-surface Latent Fingerprint Database
  • Multisensor Optical And Latent Fingerprint Database
  • Rural Indian Fingerprint Database
  • Multi-Surface Multi-Technique (MUST) Latent Fingerprint Database
  • Unconstrained Fingerphoto Database
  • Spoofed Fingerphoto Database
  • Simultaneous Latent Fingerprint(SLF) Database
  • The IIIT-D latent fingerprint database
  • NIST Special Fingerprint Databases (SD):
    https://www.nist.gov/itl/iad/image-group/resources/biometric-special-databases-and-software
    A family of fingerprint datasets from NIST used as comparison tests. Unfortunately, most of these have been withdrawn and is no longer available for purchase or download from NIST, due to the lack of the documentation required by NIST for distribution. Only the more recent SD300 series (e.g. SD301, SD302 used in NIST evaluations for latent and slap fingerprints) are currently available.
  • Hong Kong Polytechnic University Databases:
    https://www4.comp.polyu.edu.hk/~csajaykr/database.php
    The Hong Kong Polytechnic University Contactless 2D to Contact-based 2D Fingerprint Images Database
    PolyU’s fingerprint datasets are publicly released for academic research and collectively facilitate studies on hygienic contactless acquisition, cross-sensor interoperability, full ridge-surface reconstruction, and recognition under severe resolution constraints, making them versatile testbeds for next-generation fingerprint algorithms:
  • The Hong Kong Polytechnic University 3D Fingerprint Images Database Version 2.0
  • The Hong Kong Polytechnic University Low Resolution Fingerprint Database (Version 1.0)
  • The Hong Kong Polytechnic University 3D Fingerprint Images Databases
  • Tsinghua Fingerprint Databases:
    https://ivg.au.tsinghua.edu.cn/data.php

    The Tsinghua databases targets the toughest fingerprint problems: the Distorted Fingerprint Database paired normal-versus-distorted impressions for deformation studies, while the Latent Overlapped set contributes real powder-lifted overlaps and the Simulated Overlapped set adds synthetic overlaps generated from FVC2002 prints for evaluating separation algorithms:
    • The Tsinghua Distorted Fingerprint Database
    • The Tsinghua Latent Overlapped Fingerprint Database
    • The Tsinghua Simulated Overlapped Fingerprint Database
    • The Tsinghua Palmprint Database
  • Sokoto Coventry Fingerprint Dataset (SOCOFing):
    https://www.kaggle.com/datasets/ruizgara/socofing

    A public dataset of 6,000 plain fingerprint images (thumb through little finger for 600 individuals) aimed at studying fingerprint recognition and ethnicity bias (the set is predominantly African, filling a gap in typical datasets). Uniquely, it includes both original images and synthetically altered copies (to simulate challenges like scratches or obliteration).
  • LivDet Fingerprint Liveness Datasets:
    https://www.livdet.org/registration.php
    A series of datasets each containing real (live) fingerprint images and fake fingerprint images (spoofs made of materials like gelatine, silicone, etc.) captured on various sensors.