Design of Rapid Identification and Positioning Method for Disordered Safety Buckles
Aiming at the problem that the traditional production of safety buckles cannot achieve clamping with a deviation in po-sition placement,a rapid identification and positioning method for disordered safety buckles is proposed.Based on sub-templates and image pyramids,the method adopts the improved normalized cross correlation(NCC)algorithm to match the template,and combines with the circular projection algorithm to achieve the target detection of rotating the safety buckles.Meanwhile,the cross-matching and random sample consensus(RANSAC)methods are adopted to accurately locate the safety buckles.The experimental results that the compared with the actual coordinates,and the recognition error of positioning is 0～1 pixel,with an X-axis average error of only 1.02 mm and a Y-axis average error of 0.76 mm,which achieves the accurate recognition and positioning of the disordered safety buckles,and it can satisfy practical needs in engineering.
万方数据
维普
