Browse Prior Art Database

Method for Scaling Digital Images

IP.com Disclosure Number: IPCOM000036442D
Original Publication Date: 1989-Sep-01
Included in the Prior Art Database: 2005-Jan-29
Document File: 2 page(s) / 13K

Publishing Venue

IBM

Related People

Morelli, A: AUTHOR

Abstract

This article describes a method for scaling digital images of largesize technical line drawings, such as territorial maps, distribution networks, mechanical designs, etc. The method reduces the representation resolution without loosing quality (preserving the line continuity). The scaled images need less memory and transmission cost, and can be displayed on video terminals and printed on small-size paper.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 2

Method for Scaling Digital Images

This article describes a method for scaling digital images of largesize technical line drawings, such as territorial maps, distribution networks, mechanical designs, etc. The method reduces the representation resolution without loosing quality (preserving the line continuity). The scaled images need less memory and transmission cost, and can be displayed on video terminals and printed on small-size paper.

The acquisition of digital images of large-size technical line drawings (30 x 40 inches) requires high resolution scanning to capture thin lines, thus producing multi-million pixel images. Major problems are the compression of the corresponding files and the reduction of the resolution without losing the representation quality. The method addresses the above problems and includes the following steps.

The line drawing is digitized by a high resolution scanner, providing the numeric image with 256 gray levels at each pixel. A thresholding procedure is applied to obtain a corresponding binary image (black for lines, white for background).

The binary image is compressed according to a run-length-code representation. The digital image, which is a sequence of rows of pixels, is seen as a list of runs (maximal set of connected black pixels within the same row). The list is structured in blocks (containing the runs within a row) ordered according to the row sequence. each block containing the row index i and the number K:sub.i:esub. of runs within the row. They are followed by the K:sub.i:esub. couples of numbers, giving the starting and ending column index of each run. The runs are ordered according to the column sequence. Each number is represented by two bytes.

This data structure allows the handling of images with:

up to 32,767 x 32,767 pixels;

runs containing up to 32,767 pixels;

an arbitrary number of runs within a r...