Method for measuring three-dimensional coordinate by combination of laser and monocular vision in wide-angle scene

Abstract

The invention discloses a method for measuring three-dimensional coordinate by combination of laser and monocular vision in a wide-angle scene, comprising the following steps: establishing a camera coordinate system by shooting the angular point of a top left corner shot with a CCD monocular camera as an origin of an image coordinate system; placing a total station on an OB point, establishing a world coordinate system, and determining the rotation matrix and the translation matrix between the camera coordinate system and the world coordinate system; moving the total station to an OC point, and establishing a geodetic coordinate system; reading out a multi-point three-dimensional coordinate under such coordinate system, and solving a relation between the world coordinate system and the geodetic coordinate system in two positions of forward and backward movement of the total station by utilizing a seven parameter module; and reading out the image coordinate in the camera, so as to determine a three-dimensional coordinate of any relative position by using the transformational relation among the camera coordinate system, the world coordinate system and the geodetic coordinate system.According to the method, complicated steps in binocular camera calibration can be simplified and avoided, the multi-point calibration of external parameters can be implemented by only using the monocular camera and the total station, and the method belongs to the field of monocular vision and machine vision.

Description

technical field [0001] The invention relates to the fields of monocular vision and machine vision, in particular to a method for measuring three-dimensional coordinates with laser plus monocular vision in a wide-angle scene. Background technique [0002] At present, human beings usually use binocular vision to perceive three-dimensional spatial information in terms of vision, including the research on artificial intelligence and visual robots, which are more popular now. As early as the 1960s, humans used nonlinear optimization to process two images, that is, to obtain image information in different directions, and then calculate the three-dimensional pose of the spatial point. Because of its simple structure, convenient operation, and real-time operation, binocular vision is widely used in target tracking, aerospace, agricultural engineering, robot perception, and vehicle route planning. However, in the application process, due to camera calibration errors, feature extract...

AI Technical Summary

Problems solved by technology

[0003] For example, during calibration, the dual-target calibration process is complicated, and there are few controllable factors for accuracy. Especially in wide-angle scenes, the calibration is complicated, feature point matching is difficult, the line of sight is limited, and the system performance is poor.

Images