Dear users and partners!
We present the release of the VideoNet PSIM 9.1 SP6 security platform, in which the algorithms of intelligent video analytics modules –recognition of faces and vehicle license plates - have been updated.
The quality of license plate recognition has improved, new templates have been added. User interface has been improved and bugs have been fixed to improve system stability and reliability.
The following changes have occurred in the VideoNet PSIM platform:
- The Base vehicle license plate recognition algorithm has been updated. New license plate templates have been added for cars with a non-standard attachment point - square. The new algorithm allows recognition of two-line vehicle license plates.
- The Automarshal license plate recognition algorithm has been updated to version 2.38.25. Added new neural network models for Russia, CIS, Morocco, Hong Kong, Thailand. New neural network models increase the level of correct recognition of vehicle license plates in difficult optical conditions, such as low illumination and steep angles. The new algorithm allows recognition of square two-line numbers.
- New version of the Tevian facial recognition algorithm Tevian 2.12.0, in which the quality and stability of face capture and recognition have increased.
Using the face recognition module in VideoNet allows you to:
- Conduct demographic analysis
- Detect the presence of a mask on the face
- Classify the external attributes of the face (glasses, headdress, mustache, beard, etc.)
- Perform identification by face
- Identify face substitutions
- Identify facial expressions (emotions).
