Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Locating and Recognizing Barcodes on Scanned-in Forms

IP.com Disclosure Number: IPCOM000104873D
Original Publication Date: 1993-Jun-01
Included in the Prior Art Database: 2005-Mar-19
Document File: 6 page(s) / 235K

Publishing Venue

IBM

Related People

Chang, IF: AUTHOR [+2]

Abstract

A set of algorithms are disclosed that can be used to uniquely locate and identify barcodes on scanned-in paper forms or electronic forms to facilitate filing, retrieving, and form filling applications.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 29% of the total text.

Locating and Recognizing Barcodes on Scanned-in Forms

      A set of algorithms are disclosed that can be used to uniquely
locate and identify barcodes on scanned-in paper forms or electronic
forms to facilitate filing, retrieving, and form filling
applications.

      The process of converting paper forms can be confusing,
tedious, and error-prone.  One of the problems is the difficulty to
identify the forms being converted.  Usually, converted form is
obtained by scanning in its paper form using an image scanner.  The
converted form is then stored in its image format and can be
manipulated later with the help of a computer.  Due to the large
variety of the form style and the amount of different features
presented in a form, it is very difficult to uniquely identify the
form being scanned.  One of the techniques being used in solving this
problem is to use barcodes.  Conventionally, barcodes are printed on
paper form, each form or each type of forms usually has a unique code
for it.  When the form is distributed, it can be identified by using
a barcode scanner to scan through and recognize the barcode.
However, by using a barcode scanner, it is assumed that a human
operator is involved, the position of the barcode on the form is
already known, the orientation of scanning is correct, and the noise
presented on the barcode does not affect the accuracy of the scanning
process.  (If the barcode is scanned incorrectly, one just needs to
scan it again.)  Unfortunately, all of these assumptions are not true
when the form is processed by a computer.  Even though most of the
barcodes are printed horizontally and placed at the upper right
corner of a form, the position and orientation of the barcodes are
still not unique for all forms used today.  Furthermore, during the
scanning process, noise may be generated and the form can also be
distorted, which would contaminate the correctness of decoding the
barcodes.

      The purpose of this disclosure is to remove all the other
features in the form and retain the bars of the barcode, then decode
the barcode into series of 0's and 1's for final identification.  To
achieve this, four algorithms are presented, namely, the
Horizontal_Filtering (HF), Vertical_Filtering (VF), Space_Filtering
(SF), and Best_Fitting (BF), for locating and recognizing barcodes
contained in a scanned-in paper form or an electronic form.

      Horizontal_Filtering (HF)

      The purpose of this algorithm and the next algorithm is for
initial filtering of unwanted features in the form, by taking
advantage of the characteristics of barcodes.  Barcodes are
represented by vertical bars with different widths.  Although those
widths are variant and affected by the type of barcodes used, the
scaling and sizing of the form, and the resolution of the form image,
their ranges are predictable.  The HF algorithm removes all the form
features which are outside these width ranges by scanning through
each hor...