Browse Prior Art Database

Edge Detection Using Kinoforms

IP.com Disclosure Number: IPCOM000074626D
Original Publication Date: 1971-May-01
Included in the Prior Art Database: 2005-Feb-23
Document File: 2 page(s) / 38K

Publishing Venue

IBM

Related People

Hirsch, PM: AUTHOR [+4]

Abstract

Accurate placement of small parts relative to a work station is difficult to achieve, especially where the accuracy required is to within 0.001 in.

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 71% of the total text.

Page 1 of 2

Edge Detection Using Kinoforms

Accurate placement of small parts relative to a work station is difficult to achieve, especially where the accuracy required is to within 0.001 in.

Illustrated is a system for automatically positioning small parts by an optical feedback arrangement. The object to be positioned is mounted on a microscope stage and the leadscrew is driven by means, not shown, under control of a computer. Positional information is fed from the shaft encoder to the computer. A binocular microscope is used to look at the part as the part is driven past the microscope objective. Typically a 100x objective could be used, giving an aerial image at I and I'.

A kinoform high-pass filter requires encoding positive and negative coefficients of the filter operator in separate kinoforms and subtracting the results. This can be achieved using a binocular microscope, as shown.

Here the two kinoforms are placed between the aerial images and imaging lenses L and L'. The filtered real images are focused onto the masks M and M'. A slit, parallel to the edges to be detected, is placed in the center of each mask. The slit width depends upon the accuracy with which the edge is to be detected, but would generally be ,1 mil for ,20 mu inch edge detection using a 100x microscope objective.

As the piece is swept by the microscope objective the filtered images are swept past the slits in the image planes M and M'.

Behind each slit is a photodetector. The signal from the detect...