Monocular stereo vision relative position/pose measuring method

Abstract

A disclosed monocular stereo vision relative position / pose measuring method comprises: step S1, installing a monocular camera on a packaged plant, imaging a target by moving the monocular camera to acquire an image sequence containing the target; step S2, performing target detection and scene dynamic analysis on each image, if the target is mobile in the world coordinate system, segmenting each image into static background and dynamic target, respectively performing characteristic matching on the static background zone and the dynamic target zone in two adjacent images, if not, directly performing characteristic matching on two images; step S3, establishing a constrained relationship between the characteristic matching of every two adjacent images and the motion of the camera, calculating the motion parameters of the camera; and step S4, on the basis camera self motion estimation, reestablishing the real position of a 3D characteristic point in the space according the images of every two adjacent images of the 3D characteristic point, thereby calculating the relative position / pose information of the target.

Description

technical field [0001] The invention relates to the technical field of visual measurement, in particular to a monocular stereo vision relative pose measurement method for realizing the relative position and posture of the stereo vision effect of binocular cameras by using a monocular camera. Background technique [0002] With the development of computer vision theory and image acquisition equipment, measurement methods based on computer vision have been widely used in more and more fields because of their advantages of non-contact, wide measurement range, low cost, and high degree of automation. [0003] The monocular vision system uses a camera to measure the target. Since the monocular vision system loses the depth information of the target in the three-dimensional space during the measurement process, when using the monocular vision to measure the position and attitude of the target, it is necessary to have Know the physical size of the target and other information. The ...

AI Technical Summary

Problems solved by technology

[0006] Aiming at the above-mentioned deficiencies in the prior art, the present invention provides a monocular stereo vision relative pose measurement method, which solves the problem of inability to accurately measure the target depth of field and calibration in binocular stereo vision measurement in traditional monocular vision measurement Complexity, high cost, large volume, etc.

Images