Carmen MMR Software Add-on

Description
Specifications

Get geo-specific brand and model data

Carmen® MMR is an add-on product developed for the Carmen® ANPR Image, ANPR Stream, or ANPR Cloud plate recognition software. It is also available for all Adaptive Recognition traffic monitoring cameras that feature onboard ANPR.

The optional add-on enables the software to recognize the brand, model, color, and category of vehicles based on front or rear images. It brings value in many application areas, from tolling through identifying stolen vehicles to creating statistics for marketing in retail parking lots.

Carmen® MMR is available with region-specific engines. 230+ vehicle makes, 1700+ models from 7 geographical regions worldwide are recognized with high accuracy. Brand and model names are returned in a geo-specific manner.

Main benefits

  • Accurate make, model, color, and category recognition
  • Vehicle data for access control, marketing, and other purposes
  • Worldwide coverage through regional engines
  • Regularly updated database
  • Results delivered 4 times faster compared to other MMR engines on the market

Key features

  • 230+ vehicle makes (brands) recognized
  • 1700+ vehicle models recognized
  • Vehicle category recognized
  • Colors recognized even in low light conditions
  • Cloud or on-premise solution
  • 200+ makes and 1,700+ models
  • 97+% MMR accuracy (including non-ideal images)
  • License location – Cloud, On-Premise
  • Input type – Image
  • Form of software – Application, Cloud API, SDK
  • Availability – Optional to ANPR
  • Supported Operating Systems – Cloud API also available, Windows and Linux (32/64 bit)
  • Supported CPUs  – AVX is required
  • Licensing – Option to all image-based ANPR software: Carmen® ANPR Image, ANPR Cloud, Carmen® Mobile
  • Software – Requires a working Adaptive Recognition ANPR software license, SDK, Application or Cloud API
  • Recognized categories – Bus, Car, Heavy truck, Light truck, Van
  • Recognized vehicle makes – 200+
  • Recognized vehicle models – 1,700+
  • Category classification accuracy – 97+%
  • Make & model recognition accuracy – 97+% (custom made engines available)
  • Speed of recognition – At least 4 times faster than other MMR engines we tested