Sub-pixel straight lines detection for measuring through machine vision

ANA GEORGINA FLESIA; GUILLERMO AMES; GUILLERMO BERGUES; LUIS CANALI; CLEMAR SCHURRER

Montevideo

Conferencia; I2MTC 2014 - IEEE I&M INTERNATIONAL INSTRUMENTATION AND MEASUREMENT TECHNOLOGY CONFERENCE 12?15 May 2014 MONTEVIDEO, Uruguay; 2014

IEEE Instrumentation & Measurement Society

External visual interfaces for high precision measuringdevices are based on the segmentation of images of theirmeasuring reticle. In this paper, a method for subpixel straightlines detection is presented and tested on images taken fromthe reticle of a dark field autocollimator. The method has threesteps, the sharpening of the image using a version of the Savitzky-Golay filter for smoothing and differentiation, the construction ofa coarse edge image using Sobel filters, and finally, the subpixeledge location determination, by fitting a Gaussian function toorthogonal sections of the coarse edge image.We discuss results ofapplying the proposed method to images of the reticle of a Nikon6D autocollimator, using the scale of the device as a benchmarkfor testing the error in the location of the lines and comparethem with Sobel/Hough and Sobel/polynomial fitting. We reportthat for this type of image-content, Gaussian fitting has smalleruncertainty, when cameras with two different sensors are used.