RESUMEN
A novel, to the best of our knowledge, hole radius and spacing measurement method is presented based on binocular vision combined with dynamic local plane for the large-size workpiece with plane feature. The novel method overcomes the problem of the ariable working distance and large location error of an abrupt edge for conventional methods. Its measurement principle and model are introduced. Then, a chamfered inner edge extraction method based on the shrinkage of an elliptical band and an adaptive weight optimal calculation based on redundant measurement data are presented to improve the measurement accuracy. Experiments are carried out to validate the presented method, and a high measurement accuracy with a standard deviation of 0.017 mm in the hole radius and 0.021 mm in hole spacing have eventually been achieved for the actual measurement workpieces, respectively. Compared with the common method, the measurement accuracy is improved by at least 77% and 53% in radius and spacing, respectively.