When paperwork containing bar code symbols are faxed, the standard of the bar code image will degrade. Paperwork scanned by fax machines in Commonplace Decision are sampled with 204 horizontal dots per inch and 98 vertical dots per inch. Every pattern level is then transformed to both a white or black worth (a course of referred to as binarization). The binarization course of converts straight traces of the bar code symbols into ragged traces altering the width of the bars and areas. If a faxed doc is itself resent out as a fax, additional deterioration happens. Every further fax cycle continues to degrade the bar code. Sooner or later, the image could now not be capable of be decoded by bar code decoding software program.

The rest of this text will focus on the problems that must be thought-about when deciding on bar code symbologies and sizes to extend the probability of profitable decodes after a number of fax cycles.

Linear (1D) Symbologies

Two in style 1D symbologies have been examined: Code 128 and Code 39. Code 128 makes use of 4 bar and house widths to encode data, whereas Code 39 solely makes use of two. 13 module sizes (smallest component) starting from 12.5 mil to 32.5 mil of every image sort, encoding ten numeric digits, have been printed on a sheet. The sheet was despatched by way of a fax at Commonplace Decision for a complete of ten cycles. This corresponds to a horizontal pattern per module starting from 2.6 to six.6. The ensuing 11 pages have been then scanned on a flatbed scanner at 300 samples per inch and have been introduced to a bar code decode software program toolkit. At 2.6 samples per module each bar code symbologies have been solely capable of be efficiently decoded for three faxes whereas the identical codes with 6.6 samples per module have been nonetheless readable after 10 faxes.

The bigger samples per module (6 samples and better) have been capable of efficiently decode after 10 fax cycles. At 5.6 samples per module, the decode efficiency was no higher than four.6 pixels per module. This demonstrates that every fax cycle alters the bar code in a delicate means. Each fax cycle produces a novel model of the unique bar code, and every alteration could make an emblem that didn’t learn after a sure variety of fax cycles learn on the subsequent one. Nevertheless, if excessive learn charges are desired after many fax cycles, then printing the comprar container symbols at a minimal of 6 samples per module is advisable.

From the testing carried out, Code 128 barely outperformed Code 39 after a number of faxes. Given the upper knowledge density of Code 128 and the in-built checksum, Code 128 seems to be a better option between the 2 symbologies. The pictures beneath present the scale benefit of Code 128 over Code 39, with each symbols encoding 10 numeric digits with the identical module dimension. If the information to be encoded is solely numeric, the numeric compaction mode of Code 128 can be utilized to additional improve knowledge density.

There’s a commerce off to think about: the bigger the module dimension of a linear code, the extra horizontal room shall be required to encode the identical quantity of data, however the bigger modules sizes can enhance learn charge. One further issue to think about is the bar code top. For this check, all of the codes have been ½ inch in top. Given that every one vertical data in 1D bar codes is by definition redundant, if house permits, growing the peak of a bar code will usually present higher learn charges after a number of fax cycles.

Matrix (2D) Symbologies

Matrix symbologies present dramatically higher data density as a result of data is encoded in each horizontal and vertical instructions. Given their dimension benefit over their linear counterparts, 2D symbols will be printed with a lot bigger module sizes and nonetheless be akin to linear symbols encoding the identical data. For the 2D check, we printed numerous sizes of Information Matrix and Micro QR Codes. The Information Matrix was a 12 x 12 module sq. image. The Micro QR is a more room environment friendly model of a QR code that employs just one finder sample and is restricted to a lot of smaller sizes. The symbols have been encoded at a comparable degree of error correction. The module sizes diverse from 39 mil to 79 mil. Just like the 1D check, we subjected the picture to 10 fax cycles after which scanned the pictures. The pictures have been scanned at 150 samples per inch given the massive function sizes of the check bar codes. The scanned photographs have been then processed by the identical bar code decoding toolkit.

At eight.7 samples per module the 2D bar codes have been capable of be learn after 5 fax cycles and as soon as 15 samples per module have been reached the codes have been readable as much as 10 cylces.

The outcomes weren’t as constant because the 1D symbols. This is because of the truth that the distortion of the image in the course of the growing variety of fax cycles impacts knowledge in each the information dimensions. Nevertheless, we are able to draw the conclusion that symbols with bigger dimension modules will decode extra reliably after a number of fax cycles.