Vision Components' Carrida Software Engine is a powerful OEM Library, which is easy to integrate into existing security and surveillance applications. The software tool assures reliable high-speed recognition of number plates/license plates.
The software is resistant against changing and difficult lighting conditions. Furthermore, it is well suited for stationary systems and also for mobile systems when coupled with a small lightweight compact camera. The VC ALPR software (Automatic License Plate Recognition Software) can be called from any C and C++ projects via API, allowing broad range of applications.
The VC ANPR software (Automatic Number Plate Recognition Software) is quick and easy to set up and can also be trained to read codes reliably on boxes, containers or similar, even if special fonts or symbol codes have to be detected. For the result's data transfer standard network protocols are available. The recognition accuracy of more than 96% includes identifying damaged plates and angle correction.
Automatic country/state recognition with typical reading accuracy: >90%.
Vehicle plate recognition in several lanes simultaneously.
Recognition of two line vehicle plates.
Reading all common still image and movie formats from memory or file.
All plates (no limitation) displayed in one image are read.
LPR works in all countries across the world.
Special neural network technology constantly refined.
Confidence value available for each vehicle plate.
Confidence value available for each character.
Completely hardware independent.
Integration with IP, USB and GigE cameras, as well as frame grabbers for analogue cameras.
Sophisticated angle correction in two planes.
Minimum char height: 12 pixel.
Static, parking and free-flow engines available.
Live streaming of images.
Encrypted data and image transmission.
Crypto dongles or software protection used.
License required per workstation/PC unit.
System Requirements and Programming Languages
Intel x86 CPU 1.2 GHz 1 core Linux 64-bit or Windows 32/64-bit OS
ARM A8 800 MHz, NEON SIMD engine 1 core Linux 32-bit or 64-bit OS
min. 40 MB flash memory
min. 64 MB free RAM
Programming languages: C, C++, C#
Examples of ported platforms
Xilinx® ZYNQ® 7010
Samsung® Exynos® 5422
Broadcom® BMV 2837
Using Carrida on the Z-board embedded camera enabled us to implement a compact solution very quickly in a cost-efficient way. Due to our long-standing partnership with Vision Components, we know we can rely on the high quality and long-term availability of the hardware and software.