License Plate Recognition Will Find You

Ever watch an old movie where the bank robbers escape in their “get-away” car? If that movie took place today police would be able to use License Plate Recognition technology to find out that the car belonged to Clyde Barrow and they’d be nummernschild reservieren able to search DMV records for where he lived with Bonnie Parker; unless they’d stolen the car. In that case, detectives might be contacting the rightful owner.

LPR solutions are gaining popularity not only in the USA but all over the world. A major draw is its universal application since every country and state requires cars bear at least one license plate. So there is no need to attach transponders or transmitters. Cameras capture the plate and the software translates its characters into digital form so the computer can organize it for matching to whatever database the particular application calls for.

Police department have adopted LPR on red light traffic cameras,
And to reinforce compliance with speed-enforcement zones
Federal, state, and municipal governments use it to monitor and restrict automobile entry to secure facilities.
Both Highway Toll authorities and private parking facilities utilize LPR to collect toll and parking fees.
And lastly, ever see those repo men attaching a boot locking device to the wheel of a car? Know how they know it belongs to someone behind on their loan payments? Yep, LPR.

Since we’re all under LPR scrutiny, we might as well know how it works. License Plate Recognition draws on the technology underlying Optical Character Recognition. OCR uses optical scanners and specialized software to electronically recognize and digitally encode printed or handwritten characters. LPR systems depend on the following components:

IP Camera

In order for LPR to work, the image must be clear, especially when capturing fast-moving vehicles. IP cameras are the camera of choice since they offer high resolution and definition. Depending on the application, the camera is mounted so that it can capture either the front or rear license plate, and in some cases an additional camera is used to catch the driver’s face.

Illumination

Infra-red lighting has been found to work best primarily because the human eye cannot detect it. Therefor the driver will not be startled nor will any covert applications be compromised. In addition, infrared lighting works well with the retro-reflective license plates issued by most states. It is also equally effective night and day.

Frame Grabber

The Frame Grabber is the interface between the IP camera and the computer. It converts the analog signal from the camera into a digital stream of 1’s and 0’s that can be understood by the computer.

Computer

The computer is the coordinator of any LPR system. Usually a PC running Windows or Linux, it runs the LPR application and interfaces with any other systems an application might call for.

Software

LPR software can be broken down into two steps – accessing nummernschild reservieren the characters displayed on the license plate in order to recognize and extract them. To do so it uses software based on Optical Character Recognition technology to read the text of scanned images.

As mentioned above, License Plate Recognition falls into the category of Optical Character Recognition. OCR technology uses software that reads the text of scanned images. LPR software, in particular, reads manipulates the license plate image, and in a series of steps extracts the information and transforms it into a form the computer can manipulate and organize for whatever the particular application calls for.

License Plate Recognition in Action

Let’s create a scenario in which LPR is used for parking facility access control.

A vehicle approaches a lift gate and activates the system by driving over a magnetic loop.
Infra-red illumination lights up the targeted area, in most cases, the front or back of the vehicle including the license plate.
The camera captures the approaching illuminated car.
The Frame Grabber converts the cameras signal so it can be read by the computer.

Let’s take a minute to examine the Frame Grabbers actions. What it actually does is single out the area holding the license plate. It does this by employing an algorithmic function called localizing which first determines the portion of the image bearing the license plate. In order to do this, the algorithm first searches for a rectangle. Thus the first action is to disregard the headlights, grill, and mirrors. And since a vehicle may have other rectangular surfaces, additional algorithms work to find characteristics of license plates such as specific dimensions and the presence of a background color.

With the area holding the license plate isolated, the software, using its own set of algorithms, enhances the image, locates the license plate itself, isolates the string of characters, and identifies the fonts.

Again, a pause to look at these initial algorithms and how they prepare the image for processing:

One algorithm adjusts any angular skew to allow for accurate sampling, correction, and recalculation for optimal size

Another algorithm regulates contrast and brightness

Another separates the alpha-numeric characters for processing by the OCR algorithms.

The OCR algorithms now take the isolated string of characters and translate them into an alpha-numeric text entry.
And to complete the process, another set of complex algorithms verifies the alpha numeric information prescribed by the specific rule set of the software application.

It sounds like a long complex process, doesn’t it? It may be complex but believe it or not, it doesn’t take long. The algorithms run through their operations within milliseconds, with each one dependent on the preceding one.

Database

Upon completion, the resulting entries can be compared to, and/or added to a local database, or sent over a network. In the scenario we have set, entrance to a parking facility, if the sequence of characters from the license plate matches an entry on a list of authorized vehicles, a signal will be sent to the gate and it will open just long enough for the car to be admitted. If it’s not a match, the gate will remain lowered.

Uses

Other applications are meant to prevent movement. If the license plate characters on a parked car match up to an entry on a list of delinquent car loans, stolen vehicles, or traffic ticket scofflaws, the car will either be immobilized by a wheel locking device known as a boot, or be loaded on a flat-bed truck. In either case, the driver will not driving anywhere for a while.

In traffic law enforcement cases, once the license plate data has been matched up and the owner and address has been identified, a summons can be generated and mailed out to the offender. Don’t let this be you!