Image analysis – Applications – Reading bank checks
Reexamination Certificate
2000-03-03
2003-11-25
Au, Amelia M. (Department: 2623)
Image analysis
Applications
Reading bank checks
C382S213000
Reexamination Certificate
active
06654487
ABSTRACT:
REFERENCE TO COMPUTER PROGRAM LISTING APPENDIX
Incorporated herein by reference is the material contained on the compact disc filed during the prosecution of this matter. Two duplicate compact discs have been submitted, and each contains one (1) file named 09-518,424.txt and created Nov. 1, 2002. The file is an OCR and manual reproduction of the program listing filed as part of the original application. The file size is 127 kilobytes.
BACKGROUND OF THE INVENTION
This invention relates generally to computer-implemented methods for digitally recognizing a plurality of characters across an area of a digital image containing character indicia at predetermined positions in the area and also possibly containing non-character indicia in the area. The invention relates particularly to an analysis method and tool for a digitally imaged financial item, such as a check having a MICR line, and even more particularly to a check processing method and a check sorting system.
Although the present invention has broader aspects referred to above, the invention will be described in the context of handling financial items, specifically checks which have along their lower portions indicia-containing areas referred to as “MICR lines.” Referring to
FIG. 1
, a conventional check
2
has various conventional portions. These include the illustrated date, payee, numerical and alphabetical amount, memo, and signature areas. Typically other information is preprinted on the check as well (e.g., account holder identification information such as name and address, and bank name). The check
2
also includes a MICR line
4
. “MICR” stands for Magnetic Ink Character Recognition, and the acronym is used in this specification and the claims as an adjective to designate conformance or compliance with fonts and formats used in magnetic ink character recognition technology.
With regard to MICR lines and characters used on checks in the United States, the symbol set includes numerals
0
through
9
, a financial institution designation symbol
6
shown in
FIG. 1
, an individual account number designation symbol
8
shown in
FIG. 1
, an amount designation symbol
10
shown in
FIG. 2
, and a hyphen (not shown). On the check shown in
FIG. 1
, two symbols
6
define the limits of a portion
12
of the MICR line
4
within which a nine-digit financial institution identification number in MICR numerals (not shown) is preprinted on the check
2
. The symbol
8
delimits to the right (as viewed in
FIG. 1
) a portion
14
of the MICR line
4
within which a nine-digit individual account number in MICR numerals (not shown) is preprinted on the check
2
. To the right (as viewed in
FIG. 1
) of the symbol
8
is a portion
16
of the MICR line
4
in which the check number is preprinted in MICR format (not shown). During processing of the check subsequent to it being used by the account holder to draw on the holder's account, the symbols
10
and a ten-digit number in MICR format are printed in portion
18
of MICR line
4
shown in
FIG. 2
to indicate the amount of the check (the number shown illustrates certain MICR numeral forms, and it represents a check amount of $525.00).
During the time between when the account holder uses the check
2
and when it is returned to him or her as a canceled check with the account holder's monthly bank statement, for example, the check may go through various check handling processes. These can be at a bank, at a Federal Reserve facility, or at a third-party check-handling facility, for example. One of these processes will be described as one context in which the present invention has utility.
In this check processing, a check sorter transports individual checks from a batch of checks at the input of the sorter to respective receptacles, referred to as pockets, at the output of the sorter. Along the way in the sorter, the MICR line is read and a digital image of the front and possibly back of the check is made. Other functions may be performed; however, the MICR line reading and the digital imaging are of particular significance to the present invention. The electrical signals generated by reading the MICR line are conducted or otherwise transmitted to a computer programmed to determine if the read MICR line is valid or invalid, and the electrical signals representing the digitized image are conducted or otherwise transmitted to a database providing a repository of digital images of the processed checks. The foregoing is known in the art.
If the programmed computer determines that the read MICR line is valid (as defined by programmed rules), it outputs control to the sorter to put the respective check in an appropriate pocket (e.g., a pocket designated for the financial institution whose identification number is in the MICR line). If the programmed computer determines that the read MICR line is not valid, it may control the sorter to send the respective check to the “reject” pocket. Some reasons why a MICR line might be deemed invalid include: bad print of the magnetic substance to form the MICR symbols (whether the numerals or the designators), misorientation of the check in the sorter, correction tape placed over the MICR line, sorter mishandling, ferrous or magnetic quality to signature ink or other marking made on the check in the MICR line area (e.g., random marks
22
or letters
24
impinging on MICR line
26
illustrated in FIG.
3
), and the check having a reject repair strip (a reject repair strip
20
is illustrated in dot-dash lining in
FIG. 2
; this is adhered to the bottom of the main check body and it has similar MICR encoding to that found in the conventional MICR line
4
on the main check body, as known in the art). In either event (i.e., either a validated read or an invalidated read), the database is updated to include the digital image of the processed check and to include the valid or invalidated information about the check's MICR line.
Valid MICR reads usually occur in the aforementioned automated process; however, even a small percentage of invalidated MICR reads can impose a significant cost. For example, one type of check sorter may process one million items (checks) each day. A reject, or invalidated MICR line read, rate of only 1.5% results in 15,000 rejected items per day for such a sorter. Presently, these rejected items are taken from the reject pocket of the sorter and manually processed by key entry personnel, who manually handle the rejected checks, read the MICR line information, and type the numerical information into the system computer(s) to provide the correct MICR line information. A highly efficient key entry person may be able to manually process five hundred rejected checks per hour; therefore, in this example, manually processing 15,000 checks per day would require thirty hours of personnel time each day if all personnel were able to handle five hundred rejects per hour. Of course with manual entry by humans, there is the possibility of data entry error. Clearly, the manual handling of invalidated MICR line items is time-consuming and expensive. Thus, there is the need for automating the processing of checks which have been deemed to have invalid MICR lines. This includes the need for automatically correcting the invalidated MICR line information in its database. A broader need is to provide for automating character recognition (and preferably correction) of character indicia of predetermined characteristic at predetermined positions as distinguished from other markings referred to as non-character indicia (which may in fact be characters, but not of the same predetermined characteristic).
SUMMARY OF THE INVENTION
The present invention meets the aforementioned needs by providing an automated analysis method and tool for a digitally imaged financial item, such as a check having a MICR line. The present invention provides a computer-implemented method for digitally recognizing a plurality of characters across an area of a digital image containing character indicia at predetermined positions in the area and also containing non-c
Au Amelia M.
McAfee & Taft
Miller Martin
LandOfFree
Character recognition, including method and system for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Character recognition, including method and system for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Character recognition, including method and system for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3126886