ANSI Standard on Check "Image Readiness" Nears Implementation Date

The ANSI X9.7 voluntary standard on Check Backgrounds was revised in 1999 to include new "image ready" requirements for vital areas of the check. Previously X9.7 primarily addressed the Convenience Amount Recognition (CAR), which is the numerical amount area shown in Figure 1. It is critical that the banks can read the CAR rectangle and its corresponding print contrast signal (PCS) to assure the printed rectangle dropped out and did not interfere with automatic machine recognition of handwritten amounts in bank imaging equipment.

Figure 1 • "AOI" - Areas of Interest
(and CAR area)

The new standard applies to the entire face of the check. The new requirements focus on accurate scanning and binary conversion of handwriting from "areas of interest" (AOI), namely the Date, Payee, Legal Amount, and Signature areas. Although these ANSI AOIs, which are shown in Figure 1, relate to the structured personal check format, business checks, with their varied designs, are expected to conform in principle by applying the same criteria. For example, the legal amount area (amount spelled-out in words), wherever it may be on a business check, should conform to assure the expected performance from bank imaging systems. Business checks are also at issue because of their larger physical size causing huge image files if they don't conform to ANSI criteria for background clutter. Background clutter can consist of extraneous ink, offset ink that does not drop out when scanned, or any other unwanted marks. While the ANSI standard is voluntary, most banks are gearing up for conformance to ANSI X9.7 1999 by January 1, 2001, which is the date published in the standard. Many banks have published the date in check specification orders, however, the standard is voluntary, not regulation, so account holders are free to use up existing check stocks after January 1, 2001.

Background Clutter

Figure 2 • Background Clutter Measurement by Paxel Count

In order to measure handwriting scanability in these areas and sort out extraneous marks from handwritten characters, a new ANSI measuring tool was developed called a paxel. ANSI defines a paxel as "a group of black pixels in a binary image measuring .010" x .010" square, that is the smallest dark area of background clutter that has been determined to affect the legibility of handwritten data on checks. A related term, "paxel count" refers to the "number of of contiguous paxels that, when joined in any shape, line or combination can create a background clutter problem to affect the legibility of handwritten data on checks" (See Figure 2). These parameters have been specified as means to predict legibility and assure that information will be human readable in databases of stored check images. Reflectance is specified as "not less than 40%", averaging all pixels in all possible 1/8" square areas. The background clutter allowed on the AOI is specified as a "maximum paxel count of 12".

Overall image legibility has been found, in a study by an ANSI sub-group, to be jeopardized by 1) low reflectance of background and by 2) background clutter left in binary images after drop out of portions of the background design. (See the check images in Figure 3.) Low reflectance causes low contrast and unintended dropout of vital information, while high contrast background patterns cause random background clutter to remain in binary images that renders handwriting ambiguous. Background clutter is measured by creating a binary image of the AOI, then converting it from grayscale to black-and-white using an ANSI defined process, and then measuring the clusters of black pixels (paxel count) which remain after conversion. See Figure 2 for an example of paxel strings. The image-based measuring process visually shows which inks are causing low reflectance and background clutter, down to the pixel level, so that the check designer and printer can rearrange graphic features or modify the background for compliance.

Testing Devices
The new standard also specifies, in detail, the design of an image-based test device and the required image processing algorithms to perform these new ANSI measurements. This allows any company interested in testing equipment to manufacture and supply test equipment. Providing this detail results in no unfair advantages of one company over another.

Original Check

Greyscale Image

AOI Analysis

Areas of Interest Analysis

Stored Binary Image

Figure 3 • Progression of Original Check To Stored Binary
Image Check Images Copyright Clarke American Checks, Inc.
On A Lark® Design Copyright 2000 Kevin Whitlark

For testing purposes, a scanner is used to convert a check into a binary image. The test device then tests for ANSI conformance for background reflectance and paxel count in the AOIs. A display shows a binary image like the one in Figure 3, which simulates the bank image equipment, giving an indication of what a bank image of the check will look like. This allows for overall evaluation of the esthetic appeal and general usability of a binary image resulting from a particular check design. More advanced devices will calculate the image file size to indicate the storage requirements of such a design. Electronic storage size has become a major issue as banks want to store as many images as possible on a CD, or transmit them via Internet to clients.

The new standard resulted from considerable research over three years by the ANSI workgroup, which included representatives from all the key players in the check processing industry. Hundreds of check images from current industry reader-sorters were evaluated to determine just where the threshold of human legibility of handwriting lay relative to check background. As we move into the future, ANSI X9.7 (1999) will be the guide to image-ready check design. The electronic age of checks is upon us as financial institutions continue to adopt image processing to automate data entry, truncate checks, convert to image statements, replace microfilm with image archival, and move to image exchange with other banks.

Editors note:
RDM Corp is the primary supplier of equipment for this particular test. The measuring device RDM produces is called the Image Qualifier IQ-X97, and is approved by the ANSI Standard. However, for more information, contact Peter Hanna of RDM at address below.
Peter Hanna
RDM Corporation
519-746-8483 X224
peterh@rdmcorp.com
www.rdmcorp.com
The new ANSI standard can be purchased at http://web.ansi.org/public/std_info.html.

Information Source: DMIA