Browse Prior Art Database

A DENSITY PRESERVING DIGITAL IMAGE MAGNIFIER/REDUCER

IP.com Disclosure Number: IPCOM000024217D
Original Publication Date: 1979-Dec-31
Included in the Prior Art Database: 2004-Apr-02
Document File: 4 page(s) / 140K

Publishing Venue

Xerox Disclosure Journal

Abstract

A constant density algorithm which assumes that the value of a picture element (i.e, pixel) represents the average of the gray scale level of that portion (i.e., the window) of the image that immediately surrounds the pixel. The algorithm attempts to generate a new pixel value equivalent to having a different size window inversely proportional to the reduction/magnification ratio chosen.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 58% of the total text.

Page 1 of 4

(EROX DISCLOSURE JOURNAL

A DENSITY PRESERVING DIGITAL IMAGE MAGNIFIER/REDUCER
Jerry L. Potter

5,

SCANNING CRI D

Proposed Classification
U.S. CI. 358/287

Int. Cl. H04n

.64 REDUGTION C

0 0 0

FULL SIZE

a

b

156 1.56

r v \

1. 56 44 I. .12

t

FIG. I

* *

*

FIG. 2

Volume 4 Number 6 November/December 1979 805

[This page contains 1 picture or other non-text object]

Page 2 of 4

A DENSITY PRESERVING DIGITAL IMAGE MAGNIFIER/REDUCER (Cont'd)

A constant density algorithm which assumes that the value of a picture element
(i.e, pixel) represents the average of the gray scale level of that portion (i.e., the window) of the image that immediately surrounds the pixel. The algorithm attempts to generate a new pixel value equivalent to having a different size window inversely proportional to the reduction/magnification ratio chosen.

The intent of the constant density pixel by pixel algorithm is to maintain the density of a given area of the input image under reduction and magnification. The algorithm can be best understood if it is assumed that each pixel represents the collective density for the area enclosed by the grid of the raster scan. For example, in Figure la, a grid 5 is superimposed directly over the points 6 sampled by a raster scanning process. If the grid 5 is moved to the left and upward an infinitesimal amount, then each square of the grid will have one and only one scan point 6 inside it as illustrated in Figure lb.

In Figure lb, imagine that the original image lay underneath the grid. Then each square of the grid 5 would cordon off a section of the image. The constant density algorithm assumes that the pixel value represents the average of the gray scale level of that portion of the image that lies in each corresponding square. The reduction algorithm attempts to generate a result equivalent to having a different size grid superimposed over the image as in Figure lc. Figure 2 shows the relationship of Figure lb to Figure lc.

In Figure lc, the grid spacing is 1/.64 = 1.56 times as large as in lb. Thus, in Figure 2, the "grayness" of the area and the value of the first pixel of the larger grid is the sum of pixel (l,l), 56/100 of pixel (2,1), 56/100 of pixel (1,2) and 56/100 x 56/100 of pixel (2,2) where the points are coordinates of the "full size" grid. Similarly, the value of the second pixel of the larger grid is 44/100 of pixel x (2,l) + pixel (3,l) + 12/100 of pixel (4,l) + 100 x 56/100 of pixel (2,2) + 56/100 of pixel (3,2) + 56/100 x 12/100 of pixel (4,2). Using this scheme, all of the values of the reduced image can be calculate...