Go to documentation repository
Documentation for Axxon One 2.0. Documentation for other versions of Axxon One is available too.
Attention!
The processor must support the AVX2 or AVX512 instruction set (see ark.intel.com, Hardware requirements for neural analytics operation).
Face detection tool (see Category: Face) recognizes faces that are in the field of view of the cameras connected to the Axxon One Server.
If the recognized face meets the requirements (see Video stream and scene requirements for the Face detector VA and its sub-detectors), a biometric vector is created on its basis. It is stored in the Server database and used for comparison.
A biometric vector is a set of numbers of a given length that implicitly characterizes the visual identity of the owner of the biometric sample.
When you search for faces on the Client, the selected image is sent to the Server, where a biometric vector is created based on the received data.
The information about the biometric vectors stored in the database on the selected cameras for the specified time interval (see Setting a search interval) is uploaded to the RAM on the Server side and compared with the biometric vector created on the basis of the image received from the Client.
Attention!
The data in RAM is stored for about two hours. If you search for faces on the previously selected cameras and for a time interval shorter or equal to that stored in RAM, the search results will be obtained faster.
Face search helps you find in the archive:
- Similar faces (see Search for similar face).
- Recognized faces (see Search for all recognized faces).
- Face contacts (see Search for face contacts).
Face tracks are always highlighted in the surveillance window during face search.