INTLAB AUTO MMR

Vehicle make and model recognition SDK

What is Intlab Auto MMR for?

Creating control points with 24/7 optical recognition of vehicle types, makes, and models.

Intlab Auto MMR is a software development kit (SDK) for third-party integration of 24/7 optical detection and recognition of vehicle types, makes, and models, including determination of the vehicle's location within the frame, in a broad range of external conditions.

Basic specifications

       ​​up to
97%
Recognition accuracy
    5-300
ms
Single-frame recognition time
150
km/h

Maximum vehicle speed

1300
Number of supported vehicle models
>100
Number of supported vehicle makes
6
Number of supported vehicle categories

Compatibility

Windows 7
Windows 10
Windows server 2008
C/C++
C# (.Net)
Delphi

Specifications, system requirements, and the API

  • Specifications
  • System requirements
  • API

Acceptable camera tilt angles

Vehicle type recognition

Camera's vertical angle to the plane of the road  <= 45°.

Camera's horizontal angle to the direction of vehicle motion <= 360°.

Vehicle's roll angle relative to the lower edge of the frame <= 10°.


Vehicle make and model recognition

Camera's vertical angle to the plane of the road <= 45°.

Camera's horizontal angle to the direction of vehicle motion <= 60°.

Vehicle's roll angle relative to the lower edge of the frame <= 10°.

Vehicle speed up to 150 km/h
Minimum vehicle size

Vehicle type recognition

No less than 15% of the size of the frame, with resolution of at least 60 pixels per square meter.

Vehicle make and model recognition

No less than 25% of the size of the frame, with resolution of at least 100 pixels per square meter.

Supported vehicle types 

motorcycle, passenger car, van, motorhome (camper), freight transport (truck), bus

Number of vehicles simultaneously in the frame 

Image license - up to 3 vehicles;

Video license - unlimited

Number of video streams  depends on the type of license
Supported video signals individual images or live streams from an analog or digital camera
Licensing system

1 license for each instance of the recognition object, USB dongle

 

Types of licenses:

  • Image Category license (basic license) - recognition of vehicle type on individual frames received from 1 to 10 cameras, engine throughput of up to 10 frames per second, maximum of one license per 1 physical recognition server.
  • Image Category, Make & Model (basic license) - recognition of vehicle type as well as make and model on individual frames received from 1 to 10 cameras, engine throughput of up to 10 frames per second, maximum of one license per 1 physical recognition server.

Supported programming languages under Windows OS

The SDK can be used with many programming languages, like С/C++, C#, VB.Net, Java and any other languages supported native functions.
Package contents SDK distribution package, documentation, source code for examples, USB dongle

Supported operating systems

Windows 7,8.1,10 (64 bit), Windows Server 2008 R2, 2012 R2 (64 bit)

Recommended computer configuration
  • Core i3 (4th generation desktop CPU or higher) to perform recognition simultaneously on 1-2 video streams.
  • Core i5 (4th generation desktop CPU or higher, 4 cores) to perform recognition simultaneously on 3-4 video streams.
  • Core i7 (4th generation desktop CPU or higher, 8 cores) to perform recognition simultaneously on 5-8 video streams.
  • RAM: 4 Gb or greater.
  • Videocard, supporting NVIDA CUDA technology with NVIDIA CUDA с compute capability >= 3.0.
    License type Memory usage GPU memory usage Recommended GFLOPS of card for 1 recognizer instance
    Image Category 1 Gb 2 Gb 600
    Image Category, Make & Model 4 Gb 4 Gb 3000
  • Installed NVIDIA driver version 361.xx or higher/newer.

 

Engine input

  • images loaded from a file or passed in a memory buffer in RAW format
  • real-time video stream passed in a memory buffer in RAW format
Engine settings Frame resolution, rectangular recognition zone (ROI) within the frame
Engine output

Image Category license
After each frame is processed, the recognition engine outputs the following results:​

  • a set of recognition hypotheses, in which each hypothesis contains an ANSI string representation of the recognized vehicle type, a recognition confidence score (hypothesis weight), and the location of the detected vehicle within the frame;

Image Category, Make & Model license
After each frame is processed, the recognition engine outputs the following results:​

  • a set of recognition hypotheses, in which each hypothesis contains ANSI string representations of the recognized vehicle types, a recognition confidence score (hypothesis weight), and the location of the detected vehicle within the frame;

Key advantages

BROAD COVERAGE<br> 
BROAD COVERAGE
 
Highest number of supported vehicle models on the market, including more than 1,300 different models of passenger cars, buses, vans and freight transport vehicles from most global and former Soviet manufacturers. The engine also supports vehicle type recognition, including the 6 categories that represent the major types of existing vehicles.
Reliability<br> 
Reliability
 
Our vast and constantly growing test set of more than 8 million images lets us test each version thoroughly to evaluate recognition quality, ensure there are no hidden bugs, and be confident we have produced a high-quality product.
SIMPLE  <br>TO USE
SIMPLE
TO USE
Deploying and configuring the engine at a site takes little time, because there are no recognition settings. Accordingly, installing and configuring the engine require minimal qualifications.
Predictability<br> 
Predictability
 
Engine has a constant single frame recognition time which depends only on frame resolution and GPU processing power but does not depend on the number of objects in the frame and other factors.