Compatibility

This version of DetectorPack is compatible with Axxon One 1.9 Beta and later versions. 

What's new:

  1. Added the axxon-detector-pack-plugin-vi-detector_3.9.0.102_amd64.deb package for Linux OS and CPU mode, which includes a new detector—Abandoned object detection VI.
  2. Added a new detector—Privacy shield.
  3. Added new countries for License plate recognition VT 
  4. Added new countries for License plate recognition RR 
  5. Added new countries for License plate recognition IV 
  6. Introduced a cache-creating utility for neural network models.
  7. Brought back the sub-detector—Move from area to area 
  8. Supported FaceCube+ for Face detection (VL).

  9. Now it is possible to disable the collection of biometric data for Face detection, but in this case the search by faces will become unavailable.

    By default, biometric data collection is enabled.


Optimization

  1. Increased the performance of the License plate recognition RR of all versions by 3.3 times using GPU, and by 2.5 times using CPU.

    Optimization is available starting with Axxon One 2.0.

    Axxon One 2.0 is planned to be released in the near future.

  2. Increased the performance of the License plate recognition IV using CPU by 2.6 times.

    Optimization is available starting with Axxon One 2.0.

    Axxon One 2.0 is planned to be released in the near future.

  3. Now it is possible to quantize the neural network models for object types Human (top view) and Vehicle.
  4. Updated neural network model for the Similarity search detector.

Update recommendations

  1. For Linux OS: if a detector pack that includes all analytics tools is installed (not divided into plugins), then before updating, it is necessary to remove the installed package without using purge (complete package deletion), and then install the necessary plugins.
  2. To ensure the Face masking detection operation, it is necessary to install the Neuro Pack.

Update limitations

  1. If the currently installed version of the Addon Face Recognition Pack is less than 3.8.0, then after updating the Face Recognition Pack, the results of previously recognized faces will no longer be available. If you want to keep previously recognized faces, do not upgrade the analytics package.
  2. In the scan mode in Smoke/Fire/Object detection, it is necessary to use GPU for decoding.
  3. In the scan mode in Smoke/Fire detection, if you need to analyze only a part of the frame, it is highly recommended to set the detection zones as a polygon.
  4. The mechanisms for launching neural network models have changed, that is why some of the custom models (trained specifically for the object) may not work correctly. If you encounter any difficulties, please contact technical support.