Browse Prior Art Database

Area Average Image Reduction

IP.com Disclosure Number: IPCOM000036150D
Original Publication Date: 1989-Sep-01
Included in the Prior Art Database: 2005-Jan-28
Document File: 3 page(s) / 38K

Publishing Venue

IBM

Related People

Dattilo, AJ: AUTHOR

Abstract

This article describes a process for preventing the size reduction that results when printing scan information using gray scale halftones.

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 55% of the total text.

Page 1 of 3

Area Average Image Reduction

This article describes a process for preventing the size reduction that results when printing scan information using gray scale halftones.

Data obtained from scanning black and white photographs are stored as one byte per sample, representing 255 levels of gray scale. When printed, the data is converted to binary halftone. The output print size is controlled by transforming the gray scale data before the halftone thresholding is performed. This causes a size reduction when the sample frequency of the scanning device is greater than the effective line screen (smallest dot size) of the halftone.

A trivial solution to the size reduction problem is to discard samples in both the x and y directions. Although easy to implement, this approach results in severe degradation of the final print quality.

If the photograph is sampled initially at the spatial frequency to be used for final printing, better print quality is achieved. The sampling frequency of many scanners is not adjustable, but the same result can be obtained using the area average algorithm.

The gray scale data is often segmented into strips because of size considerations. The image strip consists of a number of complete contiguous scans. For example, a flat bed scanner may pass the data from the scan head to the processor through a data buffer. The device will scan into the buffer until the available buffer resource is less than one scan line. The buffer containing one image strip is then passed for processing.

The algorithm processes one image strip at a time. An input buffer containing the image strip is passed to the algorithm with the necessary parameters.

A suitable program encompassing the algorithm is shown below in pseudocode where: m = input samples per scan; n = output samples per scan; q = input strip scans; r = output strip scans;

L = i * j element input matrix;

I = k * j element intermediate matrix;

S = j element vector; i = row index; j = index for q; and k = index for n. start: a = 1/m ; b = 1/n i = 1 ; e = 0 ; f = 0

j = o to q

I(0,j) = L(0,j) ; S(j) = 0

j = j + 1

k = 0 to n...