Cognitec develops facial recognition engine for low-res images
10 February 2015 14:21 GMT

Research and development for the B9 algorithm focused on significant performance improvement for images with low resolution.

German facial recognition specialists Cognitec say a new recognition engine the firm has developed will bring a “significant” performance boost to its FaceVACS-DBScan product.

Cognitec says that research and development for the B9 algorithm focused on significant performance improvement for images with low resolution or taken in less controlled environments – often a crucial aspect for law enforcement applications.

FaceVACS-DBScan is used to compare still images from various sources to those stored in multi-million image databases and view match lists.

The application is used for ID management, allowing passport, visa and national ID agencies to compare new facial images to existing databases, find duplicates and errors, and avert identity fraud.

Also for law enforcement applications, Cognitec recently introduced a mobile version of the Examiner tool within FaceVACS-DBScan. Using the mobile tool, police officers and other agents can take photos in the field, compare them to central mug shot repositories or other databases, and instantly receive candidate lists on their mobile devices.