Method and device for positioning object

Abstract

The invention provides a method and device for positioning an object, relating to the image processing field. The method comprises steps of obtaining a global three-dimensional image of an object in a current scene and a local three-dimensional image of the object in a local scene, converting the global three-dimensional image to a global two-dimensional image, converting the local three-dimensional image to a local two-dimensional image, finding a graph constituted by all straight line segments inside the global two-dimensional image, wherein the matching degree of the line segments inside the global two-dimensional image and a plurality of line segments inside the local two-dimensional image satisfies a preset standard, and determining the position of the object in the global three-dimensional image according to the positions of the three-dimensional image corresponding to the found graph and the global three-dimensional image and the position of the object in the local two-dimensional image. Compared with other positioning methods, the object positioning method of the invention does not need any wireless signal transmitter, improves the positioning accuracy, reduces the calculation complexity and improves the independence.

Description

technical field [0001] The present invention relates to the field of image processing, and in particular, to a target positioning method and device. Background technique [0002] Robot positioning refers to the real-time determination of the position and posture of the robot itself in the environment. Robot autonomous positioning belongs to the field of robot intelligent control devices, and is the foundation and key technology in the process of autonomous navigation of mobile robots. At present, the more common robot positioning technology is mainly based on the prior environmental information, combined with the current robot position information and the information obtained by the sensor to determine the robot pose state. Positioning technology is mainly divided into absolute positioning and relative positioning; absolute positioning is commonly used for satellite GPS positioning, navigation landmarks, map matching, etc. Relative positioning determines the current pose st...

AI Technical Summary

Problems solved by technology

[0003] GPS positioning error is relatively large, and it is susceptible to electromagnetic interference; navigation landmark positioning needs to establish navigation landmarks to change the existing environment, which is time-consuming and laborious; existing map matching algorithms need to manually create environmental maps or self-created maps are not accurate enough to form a closed loop, etc. ; while odometer, inertial navigation system and other long-term or large-scale accumulative errors are large, and the positioning accuracy is poor

The current visual positioning technology mainly focuses on two-dimensional image matching and positioning, which is greatly affected by factors such as image quality and illumination, especially when there is little texture information in the environment, the error of two-dimensional matching and positioning is relatively large

There are various problems in the existing navigation target positioning methods, and it is impossible to provide pose information for the robot accurately and labor-saving

Images