of DB1, DB2, DB3 A, and DB4 are shown in Fig. 4.In live-scan fingerprints, it can be observed that the images from DBl (Lumidigm) have the best quality images highlighting the robustness of multi-spectral images.
Lumidigm Venus sensor captures the fingerprint in multiplespectrums and while fusing them, enhances the image quality.
Also, CrossMatch L-Scan Patrol has an in-built quality control mechanism and captures only those fingerprints that pass
the quality threshold. However, no such quality constraint is imposed on SecugenHamster-IV scanner, thus some of the
fingerprints in DB2 have relatively lower quality scores, as shown in Fig. 5. As expected, latent fingerprints in DB4 are
poor quality fingerprints with almost 96% of them having a quality score of 5. However, NFIQ is not designed to
evaluate the quality of latent fingerprints and a standard (open source) latent fingerprint specific assessment algorithm is
quality measure for simultaneous latent fingerprints (DB5) as well. Therefore, this is a high impact research challenge
which could be addressed using this database.