PDF417_ENCODE Program Documentation

Command Line

usage:      pdf417_enc [-t type] infile outfile [rows] [cols] [ec_level]

                  type = [pbm,ps,gif]

 ( < 30) in the output barcode.  The default if these args are not
supplied is 24 rows and 8 columns. Also if rows are supplied, then
columns must be supplied also.   If rows are set to 999, then the
program will use as many rows as needed to generate the barcode, up to
the limit of 928 total codes used.  This will allow up to 928/cols  rows for the barcode.  The last parameter is the ec_level.
This specifies the level of error correction to use.  It is optionally
supplied only if rows and columns are also supplied .   The optional
 -t type argument is used to specify the output type of the file.
 Valid types are  pbm, ps and gif.  The default type is ps (postscript).
 Setting the type to pbm gives a portable bit map output and to gif
 give a gif output.

The pdf417 program will allow data in three formats for the barcode output.
 The three are Text Compaction,  Byte Compaction, and Numeric Compaction.
 When switching between these formats (modes), the pdf417 standard uses
 a codeword to switch between these states.  The state machine for switching
 between these states is given  here.

Text Compaction

          Text compation is used to compress ascii character strings into pdf417 code words.              Charactors are assigned values 0-29 in 4 different submodes.  Alpha submode
           contains all the upper case charactors.  Lower case submode contains the lower   
           case letters.   Mixed submode contains the charactors 0-9 and some punctuation.

             Text compression is specified to the pdf417 program using the following syntax.

                                  TC  "this is a test\NL"

            This will give the string "this is a test" followed by a Carriage Return and
             Line feed.

            The text inside the quotes can be any ascii string.  To facilitate the usage of  the
            Carriage Return  (CR),  Line Feed (LF), and Newline (NL) the following should
            be used in the string:  \CR will produce a Carriage Return ( ascii 10); \LF will
            produce a Line Feed and \NL will produce a new line (  Carriage Return,
            Line Feed).
            Scanners may not scan unless they get a carraige return and line feed at the end
           of text data.   If you really want a backslash charactor in your string, you must use
           \BS in place of the backslash.

           The state transtition diagram for switching between submodes is given  here .

Byte Compaction

Byte compaction is used to encode binary information into pdf417 codewords.  With
byte compaction, up to 6 bytes of data can be encoded into 5 code words.   This is done
through a base 256 to 900 conversion.  If the 6 bytes of data are labeled b0, b1, b2, b3,
b4 and b5.  Then the base 256 number = b0 * 256^5 + b1 * 256^4 + b2 * 256^3 + b3 * 256^2
+ b4 * 256^1 + b5.    We can then find the base 900 number that corresponds to this
as   c0, c1, c2, c3 and c4 such that  c0 * 900^4 + c1 * 900 ^3 + c2 * 900^2 + c3 * 900 + c4
is equal to the base 256 above.  The c0, c1, c2, c3 and c4 values are the code words that
result.  The code words are preceeded by the Mode Latch 901 code when the number of
bytes to encode is less than 6, Mode Latch 924 is used when the number of bytes to
encode is a multiple of 6.

Byte compaction mode is specified to the pdf417_encode program using the following
syntax in the input file.

        BC "012345"
        BC "012ABCDEF00"
 

Numeric Compaction

Numeric compaction mode is used to encode base 10 numeric strings into base 900
code words.  In numeric compaction, the number string is broken up into 44 character
substrings (or less).  The 44 charactor strings are prepended with a leading 1 and
then converted to base 900.  The resulting number of codewords is always less
or equal to 15 codewords.

Numerical compaction is specified to the pdf417_encode program using the following
syntax in the input file.

    NC  "012345567"

 Node that the numbers inside the quotes must be in the range 0-9.  The string of
 numbers can be any length up to 256.
 

Error Correction Code

The calculation of the error correcting code is based on the polynomial equation
 (x -3) (x-3**1)(x-3**2) ... (x-3**n).  The coefficients of the resulting polynomial
 expansion mod 929 are then used as the weighting for calculating the error correction
 bits.   The first few polynomials in the expansion are given by.

(x-3)(x-9)           =  x^2 -12x +27

(x-3)(x-9)(x-27)     =  x^3 -39x^2 + 351 x - 729

(x-3)..(x-27)(x-81)  =  x^4 -120x^3 + 3510x^2 -29160x + 59049

(x-3)..(x-81)(x-243) =  x^5 -363x^4 + 32670x^3 -882090x^2 + 7085880x - 14348907

(x-3)..(x-243)(x-729) = x^6 - 1092x^5 + 297297x^4 - 23783760 x^3
                          + 650129490 x^2 - 5179955427 x
                          + 10460353203

A good source for finding these polynomials is at this university site.

            Number of Message Symbols           Error Correction Level           Number of Error
                                                                                                                           Code Bytes

                      1-40                                                             2                                    8
                    40-160                                                           3                                   16
                   161-320                                                          4                                   32
                   321-863                                                          5                                   64
 
 

Graphic Output Formats

The currently supported output graphics forms for the program are pbm ( portable bit map ) and postscript.

For comments or questions contact: John Lien  jtlien@charter.net
This page created with Netscape Navigator Gold