Sub-pixel level corner detection method and system

Abstract

The invention discloses a sub-pixel level corner detection method and a system. The method comprises: providing a calibration plate, and acquiring an image of the calibration plate; preprocessing theimage to obtain a first image; performing edge extraction in any four neighborhood D<ij> of the first image, and fitting four neighborhood edges in the four neighborhood D<ij> to obtain a linear equation of the four neighborhood edges; obtaining coordinates of an equidistant point O<ij> in the four neighborhood D<ij> according to the linear equation of the four neighborhood edges; obtaining a direction of the diagonal of the four neighborhood according to the linear equation of the four neighborhood edges in the four neighborhood D<ij>, wherein the diagonal is one of the two diagonals of the four neighborhood D<ij> with a lower average gray-scale value; using a SINC function gray-scale distribution to constrain a corner position to obtain a sub-pixel level corner within a domain segment ina four-neighborhood diagonal direction over the equidistant point O<ij> in the four neighborhood Dij. The sub-pixel level corner detection method reduces the influence of noise on the image acquisition of the calibration plate, and effectively improves the accuracy and the precision of sub-pixel corner detection.

Description

technical field [0001] The present invention relates to the technical field of sub-pixel level corner point detection, in particular to a sub-pixel level corner point detection method and system. Background technique [0002] With the rapid development of advanced manufacturing technology, the demand for metal parts with tiny holes in high-end equipment continues to grow. Among them, the surface quality and shape accuracy of tiny holes play a vital role in the service life and stability of metal parts. At present, the common processing route for micro-holes with a diameter of 0.1-1 mm is as follows: firstly, micro-vias are processed by EDM with micro-columnar electrodes, and then secondary processing is performed by electrolysis with side-wall insulating hollow columnar electrodes. Electrolytic secondary machining can remove burrs and thin the remelted layer, and also repair surface micro-cracks after EDM, thereby effectively improving the shape accuracy and surface quality...

AI Technical Summary

Problems solved by technology

Although the quadratic polynomial approximation method is simple and direct, insensitive to image noise, and has a low rate of missed detection and multiple detection of corner points, the detection accuracy of this method is lower than that of the gray center method and the gray center of gravity method, and it is difficult to meet the small Sub-pixel accuracy requirements in the process of micro-calibration of hole centerin

Images