Core Technologies
Fingerprint And Palm Print Identification
After 25 years of dedicated research and independent development, Hisign's fingerprint and palm print recognition technology encompasses comprehensive algorithms for fingerprint and palm print image prepro- cessing, quality assessment, image enhancement, feature extraction, rapid matching, and inspection. Particularly in the field of intelligent automatic processing and matching of latent fingerprints, continuous innovation has led to ongoing improvements in algorithm accuracy, earning national technology invention awards and Chinese patent excel- lence awards.
Hisign's fingerprint and palm print recognition algorithms are used by approximately 70% of crime investigation departments across provinces and cities in China, including for resident identity cards, electronic pass- ports, Hong Kong/Macao passes, and national fingerprint databases in Beijing, Guangdong, and others. The HAFIS system has built a database of over 500 million individuals'tenprint and has ranked at the forefront in several international fingerprint and palm print algorithm tests:
NIST ELFT: 2nd place in fingerprint accuracy
https://pages.nist.gov/elft/elft_1_x/results/
NIST PFT ：2nd place in fingerprint accuracy comprehensive indicator
https://pages.nist.gov/pft/results/pftiii/
FVC Fingerprint Verification: 1st place in fingerprint accuracy all indica- tor
https://biolab.csr.unibo.it/fvcongoing/UI/Form/
FVC Palmprint Verification: 2nd place in palm print accuracy compre- hensive indicator
https://biolab.csr.unibo.it/fvcongoing/UI/Form/
Facial Recognition
Hisign Technology's facial recognition algorithm has also undergone 25 years of research and independent development. Hisign employs deep convolutional neural networks and metric learning algorithms, trained with massive datasets, and utilizes deep learning for multi-feature fusion to achieve high-performance, high-accuracy for facial recognition. These applications include face detection tracking and positioning, attribute classification, liveness detection, deep fake detection, identity verification, and more.
A national-level facial recognition database system has been established by Hisign, capable of storing 1.5 billion individuals portrait, with a cumulative data- base size exceeding 3 billion people. In the NIST FRTE test, it ranked within the top
20% (out of a total of 555 evaluated algorithms).
https://pages.nist.gov/frvt/html/frvt11.html
Other Biometric Recognition Technologies
DNA recognition (with over 500 laboratory systems built, hosting the world's largest database capacity of 100 million individuals),
Iris recognition (which has won first place in the international CCBR algorithm test),
palm vein recognition, footprint recognition, handwriting recognition, voiceprint recognition, etc.
These are widely applied in criminal investigation, certificate management, financial services,social security, and other public safety field.
Machine Vision Technology
Industrial vision defect detection, video high-throughput computing, video target re-identification, video real-time enhancement processing, video blur target clarification, small face super-resolution reconstruction, etc.
These technologies are widely used in criminal investigation, social security, and other public safety fields.