Multi-target parallel measurement method, system and device based on binocular stereo vision and storage medium

Abstract

The invention discloses a multi-target parallel measurement method, system and device based on binocular stereo vision and a storage medium, and belongs to the field of optical large-size measurement. The method comprises the steps of adopting checkerboard calibration method for calibration, and obtaining the conversion relation between a world coordinate system and an image pixel coordinate system and distortion parameters of images; secondly, outputting a video stream, respectively capturing a pair of images at the same moment by adopting a left camera and a right camera for the output video stream, preprocessing the images, and then respectively segmenting and splicing the images to obtain a plurality of pairs of corresponding images with only coding targets; after traversing, quickly extracting angular point information, and after an initial sub-pixel point corresponding set is obtained, performing identification, decoding and matching to obtain a one-to-one corresponding sub-pixel point corresponding set; according to the optical triangle model, obtaining three-dimensional coordinates of the target points respectively, and completing multi-target parallel measurement based on binocular stereoscopic vision. According to the invention, online high-precision rapid measurement is realized, the portability is greatly improved, and the limitation of an application scene is reduced.

Description

technical field [0001] The invention belongs to the field of optical large-scale measurement, and relates to a multi-target parallel measurement method, system, equipment and storage medium based on binocular stereo vision. Background technique [0002] Binocular stereo vision measurement is a large-scale and high-precision non-contact measurement method. The essence of non-contact measurement is to obtain the coordinates of the feature points of the measured object in an established world coordinate system through various non-contact methods. Digital photogrammetry can ensure high precision while realizing non-contact measurement, and can reduce the complexity and cost of equipment brought by sensors, so it is an ideal measurement method. [0003] The binocular stereo vision measurement method uses two area array cameras to take images of the measured object, and after image feature calculation and tracking, 3D reconstruction, etc., the 3D information of the surface of the ...

AI Technical Summary

Problems solved by technology

[0004] In binocular stereo vision measurement, in order to obtain high-precision measurement results, high-pixel industrial cameras need to be used. The obtained high-pixel images will greatly reduce the image processing speed. It is difficult to meet the real-time measurement by conventional image processing methods. , rapidity requirements

When calculating the three-dimensional coordinates, it is necessary to obtain the precise projection point correspondence of a target point on the two cameras. Using conventional matching methods is more likely to cause matching errors and the corresponding accuracy of matching points is not high.

Existing cameras are usually fixed with a tripod, and cannot be moved after calibration. The portability is not high, and the usage scenarios are greatly limited.

Images