The “leaks” in our first article of 2014 are now officially confirmed, the ORPALIS MICR demo has just been released.
Wonder what MICR is all about?
Because that’s what we basically do here on this blog: we explain stuff.
Well, to jokingly put it in a nutshell MICR is an encoding technology used by Banks to check your cheques as they’re fond of funds.
But let’s better start from the very begining.
On the contrary to what one might assume, cheques aren’t a modern invention, they were widely used by the Muslims more than 1200 years ago.
From both an etymological and a functional point of view, the origin of cheques can be traced back to the Arabic “saqq“, an instruction written by a Muslim buyer ordering his bank to make a payment from his account.
This allowed Muslim merchants to safely do business at an even inter-continental scale, as they avoided the risks, difficulties and limitations implied should they have carried currency coins, gold or barter merchandise.
At that same time monetary systems in Europe were nothing but primitive if not worse, only few coins had reliable value and there were no banks whatsoever (“mais où sont les neiges d’antan ?“).
But anyways, some 800 years later when banking system started to consolidate in Europe, cheque payments started to evolve and spread and along with their fast increasing use came the problem of their clearing.
Cheque clearing consists in all operations involved since the moment a cheque is issued (the “payment promise” was commited by the payer) until the moment the payment is settled (the money reached the payee so the promise became reality).
Basically it goes like this: the payer gives the payee a cheque as a means of payment for their transaction.
The cheque itself is a special piece of paper issued by payer’s Bank to be exclusively used by that payer and, if not pre-printed, it also contains some blank fields to be filled in by the payer (like date, amount to be paid, payee’s name and payer’s handwritten signature).
The payee will then deposit the cheque at his own Bank which starts the cashing procedures.
Payee’s Bank will process that cheque by sending the cheque document back to the payer’s Bank and requesting (via inter-banks clearing system) the amount stipulated on the cheque to be payed.
Upon the request, payer’s Bank will verify the validity of the document and -if ok- will debit that amount from payer’s account (provided it has enough money) and transfer it to the payee’s bank, which will place it in payer’s account.
At this point the payer made the payment, the payer’s Bank canceled the cheque document (for example by applying a “Paid” stamp on it, as it was cleared) and placed the canceled cheque in payer’s files and, of course, the payee has the money thus ending the cheque-clearing cycle.
So payer’s Bank issued the paper cheque, gave it to the payer which gave it to the payee, who deposited it at his own Bank, which sent it back to the payer’s bank, where it has originally been issued.
Keeping in mind this travel tour, imagine that payer and payee are Americans, think that in the USA the number of active banks exceeds 6800 as of October 2013 and think that in 2006 americans wrote about 33 billion cheques.
And think that all of them need to be cleared.
Now how challenging is that?
The answer might surprise at least some of you: not really challenging.
Because in the mid-50’s, when tsunamis of cheques started flooding Banks, the Stanford Research Institute using equipment designed by General Electric Computer Laboratory created the first automated cheques processing system which is now known as MICR and they also developed the E-13B MICR font.
The technology was accepted as a standard by the American Bankers Association in 1958 and in 1963 the standard was accepted by the American National Standards Institute (ANSI).
At the same time in 1957 in France the Groupe Bull computer company developed another MICR font called CMC-7 which became standard in France since September 1964.
MICR acronym stands for Magnetic Ink Character Recognition and actually designates the entire technology not just the recognition part of it.
Basically, the technology consists of 4 main elements: (1) the magnetic ink used for printing, (2) the special fonts/set of characters, (3) the strict rules regarding character’s printing (like width and positioning) imposed by the standards and (4) the specific type of information encoded by specific groups of characters.
Magnetic ink is a key element for both automatic reading and document security.
Unlike other information encoding techniques (such as barcodes), the MICR fonts are human-readable but Bank’s specialized MICR-reading machines don’t use shape and contrast for detecting and recognizing them.
Instead, characters are recognized by their magnetic properties.
Magnetic ink (or magnetic toner) contains iron oxide and its quantity and distribution within a character, determines the character’s magnetic fingerprint which will be recognized by MICR readers.
Meaning that printing requires painfully-achieved accuracy but reading is blazing fast and reliable and it still works even if the characters were altered from a visual point of view (like pencil marks, overstamping, oily or greasy stains, etc).
Automatic reading is performed not just once but several times at extremely high speeds on a same document and if inadvertencies are encountered during the reading of the magnetic characters the cheque will be rejected.
The MICR fonts have limited sets of characters: the CMC-7 font has 15 characters (0-9 plus 5 special characters) while the E-13B has 14 characters (0-9 plus 4 special characters).
E-13B fonts are used as standards in the USA, UK, countries of the Commonwealth (such as Australia, Canada, New Zealand, South Africa, etc.) and various countries of Asia while CMC-7 font standard was adopted by most European countries, Israel, Japan and all South American countries, except Columbia.
Specifications regarding MICR printing are maintained within ANSI and ISO standards (for E-13B) or country national standards and they define all printing details such as the formation of the characters and the exact MICR line location. Characters have fixed width and have to be positioned exactly within the fixed character’s cell.
Finally, characters are grouped according to specific sequences, some being mandatory while others might not be.
A given group encodes a certain type of information such as the identification code of the Bank that issued the cheque, payer’s bank account number within that Bank or the routing transit number (RTN) which allows the automatic sorting, bundling and shipment of paper cheque document to be physically sent back to the payer’s Bank for clearance.
Let’s mention again that MICR use is not limited to cheques: MICR lines might be found on some credit cards, vouchers, airplane tickets, deposit tickets as well as in some corporate or government documents that require very fast physical sorting.
And let’s also stress out that MICR encoding was not meant exclusively for magnetic recognition, there’s a reason why fonts are human-readable too.
It doesn’t matter how the fonts are read via magnetic sensors or eye sight, “extracted” information is the same.
So organizations dealing with MICR documents needs to store these values (for archiving) along with other information contained in the scanned images.
Text, of course, is extracted to make documents searchable by Optical Character Recognition (OCR).
But there’s a problem: OCR engines usually do very well on printed characters but when it comes to MICR fonts….well…let’s just say they stumble on them.
This is why we’ve developed a MICR engine strictly specialized in MICR fonts and can be used as a complementary tool for OCR.
The ORPALIS MICR demo will show you how it works.
Should you want to share with us documents on which MICR fonts were tough to get extracted, feel free to send them to us (althought we treat any sample document we receive from our customers and users as confidential, please don’t forget to redact them beforehand but leave the MICR line intact).
Currently, the ORPALIS MICR demo provides support for CMC-7 fonts only.
But E-13B is being cooked by our Chef in our Labs as we speak and will be very soon served to you, hot.
The recipe is secret, of course, but this is normal in the art of cooking so just try our other hors d’oeuvres like PaperScan, ORPALIS PDFReducer, ORPALIS Virtual Barcode Reader Free or ORPALIS DICOM Viewer Free.
The “pièce de résistance” is the GdPicture.NET SDK, but you probably know that already.
So go ahead, don’t be shy, this bistro is always open.
UPDATE (13 January 2014)
The new version including E-13B support has just been released , click here to get it.