License Plate Recognition

March 28, 2011 by Nathan Miloszewski

I posted recently highlighting some applications of how intelligent video detects objects. One such specialized application that bears special attention and may be of particular interest to you is license plate recognition, also known as Automatic Number Plate Recognition (ANPR).

License plate recognition technology is gaining momentum with police departments to aid in identifying and capturing criminals and toll booth violaters but it alos has a variety of uses in retail, commerical, and industrial settings.

What is license plate recognition?

License plate recognition or, Automatic Number Plate Recognition (ANPR) refers to an intelligent video application’s ability to recognize letters and numbers on a license plate.

Uses

There are many uses for this technology outside of law enforcement.  Here are some examples:

Access Control

In an access control application use license plate recognition to only allow access to a facility, factory, or wharehouse to vehicles with certain license plates that you determine.

Criminal Investigation

Use this functionality to automatically search for vehicles with particular plate numbers that may have been involved in criminal activity.

Parking Lot Attendant

Intelligently track vehicles coming into a parking lot to elininate the need for expensive ticketing equipment or monitor how long cars stay in the lot to help reserve spaces for customers only and identify if a car has been abandoned.

Marketing

In a retail applications you can use this technology to identify where shoppers are coming from, what areas they shop in, and implement marketing initiatives to target customers in specific areas.

How does it work?

In a car identification example, license plate recognition is automatically performed in the following steps:

  1. Find the car:  First, identify the car in the image and use video motion detection if it’s in motion.
  2. Isolate plate number:  After the car is identified, use algorithms to pick out the plate and isolate the number from the rest of the image.
  3. Extract:  Using image analysis, extract letters and numbers from the license plate.
  4. OCR:  Apply Optical Character Recognition (OCR) transform collection of pixels into a stream of recognizable letters and numbers.
  5. Processing:  Lastly, the plate numbers and letters will then be stored in a database or compared with parameters set for your application.

Challenges

In license plate recognition applications specialized cameras are used to guarantee the best image quality possible.  This is because it can be challenging to deal with harsh conditions such as bade weather, direct sunlight, headlight glare, or simply dirty camera lenses.

License plates look different in different parts of the world too so you will want to fine tune your surveillance system to meet your needs.

Object Recognition Tool

The AXIS Object Recognition Tool helps you select Axis network cameras that meet your needs for identification of persons or objects such as license plates.

Also on that page you will find a pixel and distance calculator spreadsheet, where you can make detailed calculations to determine coverage and resolution for any Axis camera.

 

 

 


4 Comments

  • Ammanite

    Hi,
    Axis’ Object Recognition Tool and Distant/Pixel Calculator have been taken offline.

    Do you by chance have them for me to be able to download?

    Thanks.

  • Nathan Miloszewski

    Sorry I do not have a download for that.

    The best I can offer is the Axis suggestion to:

    — Read through the Identification & Recognition Tutorial

    Use the Axis Product Selector to find the best object recognizing IP cameras by clicking on the boxes for:
    —“Specify field of view”
    —“Specify resolution”

  • Awesome post, thanks for sharing.

    • admin

      Thanks!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

  • Featured Posts

  • Popular Posts

  • Read Our Feed

  • Latest

  • VoIP Post Categories

  • Archives