US 12,437,961 B2
Machine vision-based automatic focusing and automatic centering method and system
Yabin Yan, Shanghai (CN); Fuzhen Xuan, Shanghai (CN); and Ting Su, Shanghai (CN)
Assigned to East China University of Science and Technology, Shanghai (CN)
Filed by East China University of Science and Technology, Shanghai (CN)
Filed on Oct. 28, 2022, as Appl. No. 17/976,344.
Claims priority of application No. 202211007236.9 (CN), filed on Aug. 22, 2022.
Prior Publication US 2024/0062988 A1, Feb. 22, 2024
Int. Cl. H01J 37/21 (2006.01); G06T 7/70 (2017.01); H01J 37/20 (2006.01)
CPC H01J 37/21 (2013.01) [G06T 7/70 (2017.01); H01J 37/20 (2013.01)] 13 Claims
OG exemplary drawing
 
1. A machine vision-based automatic centering method, comprising:
controlling an object stage to move in an imaging distance range of an electron microscope, so as to change an object distance between the object stage and the electron microscope;
acquiring a plurality of object stage images scanned by the electron microscope when the object distance between the object stage and the electron microscope changes;
for each object stage image, calculating image definition of the object stage image according to a gray-scale value of each pixel in the object stage image;
assuming an imaging position corresponding to the object stage image with a maximum image definition among the plurality of object stage images as a target imaging position;
controlling the object stage to move to the target imaging position;
controlling the electron microscope to scan and acquire the object stage image, wherein the object stage image comprises an indenter and a sample arranged opposite to the indenter, a traveling direction of the indenter is perpendicular to a fixing plane of the sample, and an imaging direction of the electron microscope is perpendicular to the traveling direction of the indenter;
performing preprocessing on the object stage image to obtain a preprocessed image, wherein the preprocessing comprises filtering processing, noise elimination processing, and binarization processing performed in sequence;
performing connected component detection on the preprocessed image to determine an indenter area and a sample area;
selecting an arbitrary point in the object stage image as an origin O, and establishing a rectangular plane coordinate system XOY by assuming a straight line that passes through the origin O and is parallel to the traveling direction of the indenter as an X axis and a straight line that passes through the origin O and is perpendicular to the X axis as a Y axis;
respectively determining a midline of the indenter area and a midline of the sample area, wherein both the midline of the indenter area and the midline of the sample area are parallel to the X axis and perpendicular to the Y axis; and
controlling the midline of the sample area to coincide with the midline of the indenter area according to the distance between the midline of the indenter area and the midline of the sample area.