Neurodactyl has released second version of server fingerprint recognition SDK. New version includes significantly improved architecture, which simplifies integration and allows a user to implement some of the features seamlessly, without coding.
New in the version 2.0:
1) Batch mode is applied and managed automatically, when income load is increasing – an integrator shall no program batch collection, batch timeouts, etc. Batch mode feature provides faster processing and better throughput for the projects with high income load. Images and templates are processed in batches, allowing to have up to 10-100 better speed compared to one-by-one processing. Batch mode can be used for all stages of fingerprint recognition – detection, template extraction and matching, however, for detection and template extraction improvement of the throughput can be archived only on GPU, while for matching – on CPU.
2) Support of multithreading
3) Size of biometric template is 512 bytes - for serialized and in memory templates both
4) SDK size is reduced
5) Improved work with rolled fingerprints
If you want to test the Fingerprint recognition SDK, send us a request to info@neurodactyl.com