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

Line Width Measurement in Grey Level Digitized Patterns

IP.com Disclosure Number: IPCOM000050321D
Original Publication Date: 1982-Oct-01
Included in the Prior Art Database: 2005-Feb-10
Document File: 2 page(s) / 57K

Publishing Venue

IBM

Related People

Baier, H: AUTHOR [+4]

Abstract

Line width measurements without digitization errors caused by the finite size of the picture element (pixel) are obtained by normalizing and numerically integrating grey levels in the line scan pattern.

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 53% of the total text.

Page 1 of 2

Line Width Measurement in Grey Level Digitized Patterns

Line width measurements without digitization errors caused by the finite size of the picture element (pixel) are obtained by normalizing and numerically integrating grey levels in the line scan pattern.

Fig. 1A shows a line profile 1 which is obtained as an output signal when a photodetector is scanned across a line (shown schematically at reference numeral 2 with edges 3a and 3b). After digitization, the profile appears as a step-like curve 4; width w of each step corresponds to the lateral dimension of a pixel and represents the uncertainty in determining the exact position of the edge location, e.g., x(1) of edge 3a. (The edge location can be defined as point y where a 50 percent intensity increase has occurred with respect to underground signal I(min). ). This digitization error is present in all binary representations of pictures or intensity profiles, whose pixels are classified as black or white, according to a comparison of the grey value with a fixed threshold.

For obtaining measurement resolutions better than the pixel width, it is proposed to integrate all digital grey values along line profile 1 after normalization with respect to the maximum signal I(max). To suppress the influence of noise, only those values are considered for integration which have an intensity greater than I(min) + D and less than I(max) - D, where D is the maximum noise occurring in the photodetector signal.

The result of the integration represents the area under the normalized trapezoidal edge profile 5 in Fig. 1B. The exact position x(1) is then calculated by the trapezoidal rule from rectangle 6 with side length L, which has the same surfa...