Your Guide to ANPR Technology

Personalization is a key component in a restaurant’s arsenal.

It is important to make the customer feel wanted and special. There are various scalable solutions that you can use for that. At Wobot, we have some special use cases for improving customer experience via personalization.

Some of the use cases are -

• Personalized Greetings
• Serving pre-order customers
• Order forecasting

Now, you might be wondering, how can we figure out that the target customers have arrived. Well, the answer is ANPR or Automated Number Plate Recognition. We map license plates to customers & you will get notifications when the target customers arrive.

Extracting text from a number plate using a CCTV camera feed may sound like magic to some people, but it is actually possible. In this article, we will learn how exactly it is done. The underlying technology that is used to extract text is called OCR (Optical Character Recognition)

Image Conversion

OCR converts the input image into a black and white image, where the black region corresponds to the background and the white region is the text that needs to be detected.

Feature Detection

Once the region of the text is identified, then we move on to the next step, which is feature detection.

Example - Consider the alphabet ‘V’, you can intuitively identify this alphabet, but taking intuitive decisions is not a computer’s forte. So, how exactly does it identify the character? The answer is, that it learns distinct features for each character. The features, in this case, are two diagonal lines ‘\’ and ‘/’.

Code Conversion

There is one last step before the computer can actually identify the character: code conversion. Till now, you have extracted the text from the number plate, and let’s say the first character is a ‘V’.

Now, for you it is very easy to understand that the alphabet is ‘V’, but the computer does not understand plain English. So, we need to translate it into a language that the computer can understand. A computer can only understand binary code.

It might sound surprising to you, but the computer only understands two characters, 0 and 1. The code conversion of the character ‘V’ in binary is 01010110. This binary value here is called the ASCII value. ASCII stands for American Standard Code for Information Interchange. It simply defines a standard of how these characters should be interpreted, so any computer present in any location of the world will decode the character ‘V’ as 01010110 only. So, there is absolutely no ambiguity.

In this way. In this way, we loop over all the characters of the license plate. Once we have all the numbers, we can map the license plate to a unique customer and send a notification once that user arrives.

We have a lot of such cool use cases that can help you improve your customer experience. You can head over & schedule a demo or simply start with a free trial and see our use cases in action.