Let's you use an Optical Character Recognition (OCR) AI to read number plates in parallel or as an alternative to scanning IDs.
For events like marathons and mountain bike races it is impractical to hand-out or scan IDs, but we can use OCR to read the number plates of the runners/riders and send them a link to their gallery of images. OCR is enabled and configured at the Event level in the ELEMENTICE™ Console.
There are two main things that you will most likely need to use in conjunction with OCR for a good user experience:
Attendees - a pre-populated list of attendees/participants and some form of contact for them eg cell/mobile phone number which is uploaded to the ELEMENTICE™ Console before the event.
A Regex - a dataset used to let know the OCR what to expect, also used to check/improve the results of OCR
Once configured correctly, you should be able to photograph a runner/rider as they pass you (with a visible/readable number plate), the images will upload and that runner/rider will receive an sms message or email with a link to their gallery.
This is not a simple turn on/turn off like most of our features. If you wish to use this feature, please read this User Guide and then contact ELEMENTICE™ (sooner rather than later) for help, to make sure you have everything you need for this to work in your situation.
Enabling Optical Character Recognition (OCR)
OCR is enabled per event via the Event Page in the ELEMENTICE™ Console.
Open the ELEMENTICE™ console and click on "Events" tab on the top banner of the Console and select the relevant Event.
You are now on the Event Page and can access the Event Settings section at the bottom of the Page. "Use OCR" is one of the configurable settings under Event Settings.
Once you enable Use OCR, a new setting will appear titled OCR Regex.
By default, our OCR is set to look for any number between 2 digits long and 5 digits long.
Without a regex: If your event has number plates which are 5 digits long, but the photograph only shows the last two digits of somebody's number plate (eg xxx25) the system will read 25 and add it to the gallery of attendee number 25.
With a regex: We can set a minimum number of characters that need to be detecting before a match is confirmed, or maybe the number plate also uses numbers and we need to let the OCR know to expect that...
...basically, it's complicated, but a Regex makes everything better - only if done right. Please contact us for help building a regex for your event.
So far, we've configured the event to be able to use OCR to recognise number plates, and the console will be able to sort the images into galleries, but we still need a way to tell the individual how to access their gallery of photos.
*OCR is AI-based and thus not 100% accurate. To give the OCR AI the best chance of accuracy we will give you recommendations to make sure that your images are going to be of a sufficient resolution and your photographers are mindful of their positioning, so that the participant's number plates are a reasonable size and fully visible - which will allow the recognition to make reliable matches.
*Once Beta is finished, there will be a tiny fee per image processed with OCR. See here for more on fees & costs.