Carrida Technologies updates Edge software and SDK engine


Carrida Technologies has released version 4.8 of its Carrida Edge LPR software and an extensive update to the Carrida SDK engine.

Carrida Edge now supports MQTT to publish reading results and images enabling integration with IoT devices and networks. The hashing of license plate strings for fully anonymized storage and transmission is now supported in Carrida SDK 4.7. This makes the software suitable for set up solutions with LPR-based travel time measurement for smart mobility, traffic management, tolling, enforcement and many other data-driven applications.

The LPR engine can be flexibly deployed as OEM library, web solution for easy setup and configuration of edge devices, or with the Carrida App for any Android device. Carrida also offers validated platforms and components for easy, quick and cost-effective developments of individual LPR applications.

With Carrida Edge 4.8, MQTT publishing features have been added and enable the configuration of one or more MQTT servers. Devices using the software engine can be easily integrated in IoT networks and can securely transfer data to other devices. Data security has also been improved with public key authentication for FTP actions and certificate based authentication for HTTP data transfer. The LPR server is now even more robust and reliable and the Video4Linux (v4l2) camera handling has also been improved.

The WebGUI for easy setup of edge devices comes with enhanced features in Carrida Edge 4.8. Configuration of parameters is now much more efficient and faster, and the design of result visualizations has been improved.

Carrida Edge is based on the Carrida SDK engine, which is now released in version 4.7. This update also provides major changes, such as new license plate verification features for a better suppression of unwanted readings. Added parameters for state recognition improve reading results and enable even faster identification. The Carrida engine also now recognizes if vehicles are approaching, leaving or static, based on changes in size of the detected license plate.

Carrida Edge can optionally use securely hashed strings to transmit and store license plates, guaranteeing full anonymity of license plate readings. Combined with MQTT data transmission, applications can measure time of travel for vehicles passing two or more cameras, all in compliance with the highest data security standards.

Carrida Edge runs hardware independent on any edge device. The company offers a number of validated platforms, from Raspberry Pi to latest NVIDIA processors. For plug-and-play installations the Carrida Cam is available as ready-to-use device or hardware kit for individual designs. It has been designed for stand-alone outdoor applications and offers best reading rates, even with difficult lighting situations.

Measuring only 91 x 62 mm, the camera can be easily integrated into existing systems for stationary as well as mobile applications. It is equipped with a Sony image sensor that ensures best possible image quality and first-class reading results, even under difficult lighting and weather conditions and with changing irradiation. The ultra-compact ANPR camera is protected against dirt, dust and moisture in accordance with IP67 and thus suitable for permanent outdoor use, also thanks to the integrated heater.

Carrida Cam’s reliability and reading accuracy is based on the integrated Carrida SDK ANPR-Engine. The versatile library recognizes license plates from all countries worldwide and also accepts special formats such as two-line number plates as well as damaged or bent license plates. The software runs hardware independent on any edge device and can also be used with traditional IP-, USB- or GigE-Cameras. Its fast processing time enables instant access to the car parks. In addition to parking applications, the Carrida SDK can also be deployed for other uses in traffic management, tolling, law-enforcement and smart cities. Depending on the hardware, it can also provide make and model recognition and further data, powered by the latest Deep Learning functionalities.

Share this story:

About Author