Detection of Calibration Pattern using Concentric Circles and Bimodal Distribution of Intensity Histogram

摘要

A calibration pattern detection method is proposedrnwherein a bimodal distribution is approximated for thernintensity histogram of the calibration pattern. Because thernhistogram of the calibration pattern, such as black and whiternsquares, is assumed to be bimodal, pattern detection isrnemployed to determine the best threshold separating the twornmodes of the histogram. As most real images may includernuneven illumination, the method divides the original imagerninto subimages and then utilizes a different threshold tornsegment each subimage and approximate image parts ofrnwell separated modes. For binary images obtained fromrndifferent threshold values of different subimages, regionrnlabeling and boundary tracing of labeled regions arernperformed and ellipses with close centers are detected asrnreference points of the calibration pattern among manyrnbackground clutters.