Page History
...
- The Archive search channel module is used for searching the captured face in the face database and viewing the video archive of the captured face.
- The Age/gender defining channel module is used for age and gender recognition via the specified channel.
- The Visitor counting channel module is used for counting visitors, repeated passes, etc.
The Recognition channel and Tevian/Cognitec/VisionLabs/HUAWEI recognition module are used for real-time face recognition via the specified channel.
Note title Important! For correct operation of the recognition modules that run only on the x64 platform, the Run x64 modules checkbox should be checked on the settings panel of the Computer object (see Face recognition modules).
To activate the Age/gender defining channel, Archive search channel, Recognition channel and Visitor counting channel modules, select the required Face detection object on the settings panel of the corresponding object (see Configuring the captured faces recognition parameters).
| Info | ||
|---|---|---|
| ||
If the Age/gender defining channel determines the age of the same person with poor accuracy, then it is necessary to set the 1 or 5 value for the Tevian.TrackBestFramesCount key (for details, see Registry keys reference guide, for information about working with the registry, see Working with Windows OS registry). |
Activation of the Cognitec recognition module
...
The activation of the Tevian capture and recognition module is performed using the intellect.sec key file. Additional key files are not required.
The module is activated by the standard activation procedure that includes creating a Tevian recognition module object on the basis of the Face recognition server. A child object that shows the number of faces in the database should be created on the basis of the created object:
| Info | ||
|---|---|---|
| ||
Standard memory consumption for the Tevian recognition module is up to 500 Mb per channel. For small databases (up to 1000 faces), the module can handle events at a speed of no more than 1500 events per second. For larger databases, the number of compared faces is from 390 to 550 per second for one processor core. |
...