Hello,
Happy release day for version 2022.2, a version filled with one new and extremely handy key feature, nice enhancements for our ANPR solution and several general improvements.
Let’s take a closer look at what you can expect this time!
New Key Feature
Rule Engine: Support for custom event workflows
It’s all about customization! Now you can use the new rule engine for your custom events and workflows in your individual configuration. This means you can easily receive alerts for wrong-way driving, speeding, U-turns ad even crosswalk safety incidents!
The feature finally allows it to gather relevant data instead of Big Data. If you are keen about counting cyclists on a cycling path and you are not interested in the other traffic around this area, you can customize this with the rule engine. Have a look at it in this short video clip:
The number of vulnerable road users is increasing in cities and residential areas. The rule engine can be used to monitor crosswalks for safety. Get the information when and how often it happens that vehicles pass crosswalks at the same time as pedestrians.
New Features
Enhancements and country meta information for ANPR
Good news for all of you using our ANPR solution: We have worked extensively on ANPR to improve accuracy. As a result, we can offer a solid ANPR solution up to 40km/h movement speed with over 90% accuracy in challenging outdoor scenes at entries and exits without any special deceleration installations:
On top of that, we have added more meta information for license plate recognition: It will be possible to see the country of the license plate. This will allow you to get more information about your parking visitors, whether it’s for payment solutions, enforcement, or for identifying the region the parking customers are coming from. This country classification works with an excellent accuracy of 99%.
Improvements
Traffic model improvements
We can happily announce that we have boosted our accuracy by another 3-5% for people and bicycle detection in our Traffic and Parking models with better detections and more stable tracking. In addition, we have focused on excelling in sunny, bright conditions where shadows caused issues in the past.
License plate captures via MQTT
From now on, there is the possibility to transmit the license plate capture images within a MQTT message. This image is encoded with BASE64. If you are interested in this feature, please contact our support in order to enable it in your solution.
We hope you are as excited as we are about this release!
As usual, we offered an online live demo for this release. Feel free to check out the recording of it here.
If you have any questions, please feel free to get in touch with us!
Best regards from Innsbruck,
your Swarm Analytics team
