Browse Prior Art Database

Document Image Processing Method

IP.com Disclosure Number: IPCOM000104227D
Original Publication Date: 1993-Mar-01
Included in the Prior Art Database: 2005-Mar-18
Document File: 2 page(s) / 44K

Publishing Venue

IBM

Related People

Amano, T: AUTHOR

Abstract

Described is a method for processing document images, which smears black pixel components, leaving characters separated from figures and from characters in different columns. Incorrect concatenations due to the smearing process can be prevented by using field-separator information. A field-separator herein means a vertically long white region located on the border of a text column area and a figure area.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 78% of the total text.

Document Image Processing Method

      Described is a method for processing document images, which
smears black pixel components, leaving characters separated from
figures and from characters in different columns.  Incorrect
concatenations due to the smearing process can be prevented by using
field-separator information.  A field-separator herein means a
vertically long white region located on the border of a text column
area and a figure area.

      The figure shows the configuration of this invention.  It
consists of two image data memories used for input and output
respectively, a field-separator detector, a field-separator recorder,
and a run-length-smearing processor.  A field-separator detector
first accesses image memory 1 to detect field-separators.  It does
not have to detect field- separators specifically.  A simple
implementation of the detector is to scan an image in the vertical
direction, and detect long white pixel runs.  All possible candidates
are recorded in a field-separator recorder.  A run-length-smearing
processor raster-scans the data in image memory  1, and replaces
short (less than an appropriate threshold value) horizontal white
runs with black ones (smearing).  Resultant smeared image data are
stored in image memory 2.  The run-length-smearing processor uses two
threshold values based on the information obtained from the
field-separator recorder.  An ordinary threshold value corresponding
to a single threshold in conventional ...