Scaling techniques for printing bar code symbols

Details Scaling techniques for printing bar code symbols Scaling techniques for printing bar code symbols

2001-12-10

2004-04-27

Le, Thien M. (Department: 2876)

Registers

Coded record sensors

Particular sensor structure

C235S462010, C235S462080, C235S462160

Reexamination Certificate

active

06726104

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates to a method for representing a digital image of a bar code or matrix code in a vector format, that automatically adjusts its critical dimensions to correctly match the dot spacings of different printers and that allows various graphical parameters of the bar code image to be safely changed after the image is encoded.
BACKGROUND OF THE INVENTION
Ever since the 1960's, and especially since the advent of the Universal Product Code in the 1970's, bar code symbols have been widely adopted as the most popular form of optical marking for machine recognition of printed information. It has proved to be substantially more reliable than Optical Character Recognition (OCR) technology, primarily because (unlike printed letters and numbers) the bar-and-space “languages” represented in bar codes were designed “from the ground up” to be easily discriminated by machine, even in the presence of some degree of printing and scanning defects. Several barcode “languages,” called “symbologies” have been developed, each with its advantages for certain applications. One characteristic that they all have in common, is that the bars and spaces are varied in their widths, to form predefined patterns that actually carry the information. Moreover, these widths are not continuously variable (like the set of real numbers). Instead, they are quantized into a small set of allowed nominal widths, measurably different from each other, so that the reading machine (i.e. scanner) can pick the nearest ideal width which is the best fit to the imprecise signals that are received when light is reflected from an imperfectly-printed symbol. It is very important that the reading system can unequivocally determine which widths were intended to be printed. A single mistake in this regard will result in failure to decode the barcode, and worse, multiple mistakes can result in an erroneous interpretation of the encoded data.
Although not all bar code symbologies use the same set of allowed widths (nor would that restriction apply to the current invention), it is simplest for explanatory purposes to consider a typical class of bar code symbologies called (n,k) codes, where the bar code is typically constructed of a series of patterns, each pattern usually representing one or two data characters, and each pattern made up of a small number of bars and spaces of predefined relative widths. For example, in barcodes of the Code 128 symbology, each of these bars and each of these spaces must be chosen from one of four relative widths. The unit width (by definition, the narrowest bar or space allowed) is called a module. All of the bars and all of the spaces within a Code 128 symbol are nominally either 1, 2, 3, or 4 modules wide; no other widths are defined. The absolute width of a module is not a fixed dimension across all Code 128 symbols, but can be chosen at the time of printing. The typical width of a printed module ranges from 0.005 inch to 0.040 inch (but is fixed at one size, within any one printed symbol). Other symbologies, such as Code 39, utilize only two different allowed widths (“wide” and “narrow”) for the bars and spaces. The wide elements are not necessarily an integer multiple of the narrow elements (for example, the wide bars and spaces can be chosen to be 2.5 times the width of the narrow bars and spaces). One skilled in the art would recognize that such symbologies can also benefit from the techniques of the present invention.
Since the typical bar code reading device can operate over a wide range of distances from the bar coded paper or package, the decode algorithms in the reader do not attempt to estimate the actual sizes of the bars and spaces (they will appear smaller, the further the reader gets from the paper). Instead, the decode algorithms attempt to judge the relative (not absolute) widths of these elements. As an example, the Code 128 symbology uses a set of 108 different patterns, and each pattern is comprised of three bars and three interleaved spaces which always have a total width (across those six elements) of eleven modules. Thus, the decode algorithm (in greatly simplified form) can look at a group of three bar and three space measurements from the scanner, calculate that total width, assume that this represents eleven modules, and then judge whether each bar or space is closest to {fraction (1/11)}, {fraction (2/11)}, {fraction (3/11)}, or {fraction (4/11)} of that total width. In actuality, the decode algorithm estimates the widths of adjacent bar/space pairs, rather than single bars or spaces, because this approach is more immune to a phenomenon called uniform ink spread. It can be seen that the actual size of a printed module is irrelevant to the decoder, but the relative widths must be fairly close to the ideal ratios of 1:2:3:4. Some error is tolerated, because if a bar whose nominal width was 3 modules is accidentally printed as 3.25 modules, the decoder will still correctly round-off to the nearest allowed width of 3. Should that error exceed half a module, on the other hand, a reading failure will occur. This error tolerance must be shared between printing inaccuracies and scanning signal inaccuracies (the scanner may be too far away and somewhat out of focus, for example). Thus, the more accurately the relative widths of the bar code are printed, the more likely the system will still work despite a poor scanner signal (due to operator error, electrical or optical noise, or other factors).
Naturally, it is usually desired for bar codes to be printed in as small an area as is possible (for instance, to leave more room for other graphics and text). This is done by printing at the smallest nominal module size that is still within the capabilities of the printer to print accurately, and within the capabilities of the scanner to scan accurately. From the printing perspective, the lower limit on printed module size is somewhat dependent on the roughness of the paper and characteristics of the ink, but it depends primarily on the dot spacing of the printing device. The narrowest bar that any printer can print is, by definition, one dot wide, but the width of a single column of dots varies too much to be an accurate unit width. The standard rule of thumb is that the unit module width must be at least two dots wide, and a minimum of three dots wide is recommended. Since different printers have different dot spaces (commonly measured in dots per inch or DPI), a three-dot-wide unit module will be different physical sizes on different printers. The printing technologies that print bar codes vary by more than an order of magnitude in their DPI resolution, from 200 DPI (inexpensive thermal printers that generate small labels) to 4000 DPI (high-resolution printing presses for magazines and other high-quality print jobs).
The foregoing discussion indicates why, in the current state of the art, it has always been important to ascertain the intended printing resolution before generating a digital bar code image. This is because no printer can place a dot half way between the imaginary grid points that define that printer's dot pitch. If, for example, a bar code will be printed on a 300 DPI printer (typical of many ink jet and laser printers), a three-dot-wide bar will be nominally {fraction (3/300)} of an inch wide (i.e., {fraction (1/100)}, or 0.010 inch). A four-dot-wide bar will be 0.0133 inch wide, and a five-dot-wide bar will be 0.0167 inch wide. Note, however, that a 0.015 inch width cannot be accurately printed on a 300 dpi printer, because that width “falls between the dots” of the printer's imaginary grid. A 400 dpi printer can accurately print that width in 6 dots, and a 600 dpi printer can do it accurately in 9 dots, but the image of a barcode with a 0.015 module width will inevitably be rendered quite inaccurately if sent to a 300 dpi printer. In fact, the widths of the narrow bars will (usually randomly) be a mix of 0.0133 and 0.0167 inch wide (but none can be 0.015), and the larger bars and spaces will be similarly “

Affiliated with

Also associated with

Rating

Scaling techniques for printing bar code symbols does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Scaling techniques for printing bar code symbols, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Scaling techniques for printing bar code symbols will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3218971