A monocular stereo vision measurement method

Abstract

Disclosed is a monocular stereoscopic vision measuring method. Visual image collection is achieved by moving a coordinate measuring machine through a video camera. A measured workpiece is placed in a measuring position on a workbench of the coordinate measuring machine, the coordinate measuring machine is moved to a position A in the X direction, a photo is shot by the video camera, afterwards, the coordinate measuring machine is moved to a position B in the X direction, another photo is further shot by the video camera, the two photos are transmitted to a computer which is provided with identification software through a data line of the video camera to conduct image processing, then feature points are obtained, matching the feature points of the two images is achieved by using a feature point matching algorithm based on translation, space coordinates of various feature points are respectively calculated by the coordinate measuring machine, and the pose parameter of the measured workpiece is solved by further combining corresponding coordinates of the various feature points in a machine coordinate system and a part CAD coordinate system. According to the monocular stereoscopic vision measuring method, the coordinate measuring machine is used for achieving single video camera stereoscopic vision measuring, and three-dimensional coordinates of the part features points are obtained rapidly and accurately.

Description

technical field [0001] The invention relates to a visual measuring method, in particular to a monocular stereoscopic visual measuring method driven by a three-coordinate measuring machine. Background technique [0002] The three-coordinate measuring machine is a standard three-dimensional precision measuring equipment in the manufacturing industry. Its own automation and intelligence are also continuously improving with the development of modern industry. The intelligent three-coordinate measuring machine has become a major research in the field of measurement at home and abroad. hotspot. The automatic identification function of the part pose is an important symbol of the intelligence of the three-coordinate measuring machine. Quickly obtain the position and orientation information of the measured part, guide the measuring machine probe to automatically find the measured part, and provide a reference for establishing an accurate part coordinate system, which is the first st...

AI Technical Summary

Problems solved by technology

At present, the bottleneck encountered in the improvement of stereo vision measurement accuracy is mainly the relatively low relative pose calibration accuracy of the two shooting positions, which makes it difficult to improve the accuracy of existing stereo vision measurement methods

For monocular measurement, whether the existing technology adopts manual or electric servo drive, the positioning accuracy is low and the translation between the camera and the measured part cannot be guaranteed, resulting in low relative pose calibration accuracy of the two shooting positions and two The image feature point matching algorithm is complex, resulting in relatively low stereo vision imaging accuracy

Images