System calibration method based on depth information

Abstract

The invention relates to a system calibration method based on depth information. The method includes: firstly utilizing a circular plane target, extracting a feature point sub pixel outline by adopting self-adaptive threshold value edge detection and a multinomial fitting algorithm, and obtaining a central coordinate by using ellipse fitting; working out the optimal solution of a camera parameter by using a least square method according to a nonlinear imaging geometric model with distortion to obtain the pose of the target; at least providing depth information between the surface of an object to be detected and the target plane as the correction item of the camera model for correcting the pose of the measurement plane, and determining parts dimension by using an imaging principle and a point of intersection of a straight line with the parts surface; according to the system calibration method based on depth information, the problem that the dimension measuring system for the large-dimension plane parts is not high in calibration precision, thus being worthy of popularization and application.

Description

technical field [0001] The present invention relates to machine vision measurement technology, and in particular to a system calibration method based on depth information. The attitude transformation of the plane provides prior information, belongs to the field of machine vision, and can be applied to system calibration when measuring objects with different thicknesses. Background technique [0002] Machine vision measurement technology has the advantages of non-contact, high precision, and easy automation. Among them, the monocular vision measurement system has a simple structure and easy operation, and has gradually become an important measurement method in the field of high-precision measurement of part dimensions. [0003] The key to high-precision dimensional measurement is camera calibration. At present, the calibration method of small-sized thin slice parts can achieve high accuracy, but due to the influence of the field environment, light source, system resolution, e...

AI Technical Summary

Problems solved by technology

[0003] The key to high-precision dimensional measurement is camera calibration. At present, the calibration method of small-sized thin slice parts can achieve high accuracy, but due to the influence of the field environment, light source, system resolution, etc., the high-precision calibration method of large-sized objects is not high. Directly affect the dimensional measurement accuracy

First of all, the large-size target image has large distortion, which requires a suitable target feature point extraction algorithm

Secondly, the change of part thickness makes the measurement pose inconsistent with the calibration pose, which affects the measurement accuracy

[0004] The existing typical camera calibration algorithm is the calibration method based on the plane template proposed by Zhang Zhengyou, which has high precision and does not need the motion information of the template, but this method needs to manually extract the image corners for matching, which is not conducive to the re

Images