Metrici CCR engine
Distributed container-code recognition (CCR) engine
Features
- Architecture
- Distributed detection and recognition, free-flow or triggered
- Recognized standard
- ISO 6346 container codes — owner code, equipment category identifier, serial number and check digit, with the size/type code
- Working mode
- Continuous (around the clock) or on external/virtual trigger, or a combination of the two
- Minimum code width
- 200 pixels recommended for reliable recognition
- Number of recognized container codes in each frame
- Unlimited
- Check-digit validation
- Yes — ISO 6346 modulo-11 check digit used to confirm each code was read correctly
- Repeat detection filtering
- Yes — configurable "same code" delay to suppress duplicate events, plus a single-CCR mode
- Trigger type
- Over UDP from a Barix Barionet 50 LAN controller, from the Metrici CCR web interface, or from another Metrici engine (LPR, PPD, QR, etc.)
- Chained trigger
- Yes, to another Metrici engine (LPR, PPD, QR, etc.)
- Companion stream
- Yes
- Container tracking
- Yes — direction of movement (coming or leaving) is determined by analyzing successive image frames
- Detection window
- Yes, user definable
- Single CCR mode
- Yes
- Barrier and traffic light control
- Yes — open/close barrier and traffic light driven over HTTP, depending on the recognized container code
- GPS coordinates
- Yes, if a GPS device is connected
- Weight
- Yes — the weighing scale should be connected to a Barix Barionet 50 equipped with special Metrici firmware
- Speed
- Yes, by using a Metrici Observer Radar
- Data type for each recognized container code
- Container code, moving direction of the container, first seen, last seen, location name, camera name, GPS latitude, GPS longitude, container picture, container-code picture, companion picture, weight, speed
- Parallel processing
- Adaptive multithreading, up to 8 threads for each connected camera
- GPU acceleration
- Yes — Intel, AMD or Nvidia supported
- Data push
- POST method over HTTP; two events generated: check_action and reporting
System requirements
- OS
- Ubuntu 24.04 LTS MATE, Linux CentOS 7 (64-bit)
- CPU
- Intel i5, i7, i9 or Xeon; AMD Ryzen, Threadripper or Epyc
- GPU
- Intel, AMD or Nvidia; mandatory for more than 2 cameras/server
- System memory
- 512 MB for each connected camera, but no less than 4 GB/server
- GPU memory
- At least 256 MB available for each connected camera
- Storage
- Up to 512 KB for each container code stored in the buffer, depending on the camera resolution and image compression
- Cameras/server
- Up to 256 connected cameras on each server; unlimited number of servers
- Camera type
- IP; MJPEG, MxPEG, H.264 and H.265 video stream; HTTP or RTSP/RTP transport protocols
- Supported camera producers
- ACTi, Arecont, Avigilon, Axis, Bosch, Dahua, Diviotec, Hikvision, Mobotix, Novus, Pelco, Samsung, Sony, Uniview, Vivotek