Automatic mapping robot mapping and finishing method based on visual marks

Abstract

The invention provides an automatic mapping robot mapping and finishing method based on visual marks and belongs to the technical field of navigation. The method comprises the following steps: (1) detecting visual marks in a to-be-detected environment; reading localization information of corresponding visual marking points contained in the visual marks; based on angle point information of the visual marks, calculating the relative position and course of the automatic mapping robot relative to the visual marking points at the moment; and (2) obtaining the coordinates and the course of the automatic mapping robot according to the localization information of the visual marking points and the relative position of the automatic mapping robot relative to the visual marking points, and correctingand optimizing the mapping data having deviation and errors between the visual mark points. The method is low in cost and easy to implement, and can realize real-time accurate correction of the localization errors in the high-precision map acquisition process and real-time finishing and off-line optimization of the acquired map information, thereby omitting the requirement for loop detection in atraditional simultaneous localization and mapping process.

Description

technical field [0001] The invention belongs to the technical field of navigation, and relates to a method for constructing and trimming a map, in particular to a method for constructing and trimming a map by an automatic map-building robot. Background technique [0002] In recent years, with the gradual application of self-driving cars in some fields, the demand for high-precision map collection is increasing. In many indoor and outdoor scenarios, the GPS signal is weak, the installation cost of WiFi and UWB is high, and the positioning accuracy is not high, which makes the built map shift and distort. Constrained by the problems of drift, cumulative error, and high price in the current positioning scheme, the methods such as loopback detection and map matching adopted by traditional error elimination technologies have certain requirements on computing power and environmental characteristics. The demand for accurate positioning and error correction makes the image acquisit...

AI Technical Summary

Problems solved by technology

In many indoor and outdoor scenarios, the GPS signal is weak, the installation cost of WiFi and UWB is high, and the positioning accuracy is not high, which makes the built map shift and distort

Constrained by the problems of drift, cumulative error, and high price in the current positioning scheme, the methods such as loopback detection and map matching adopted by traditional error elimination technologies have certain requirements on computing power and environmental characteristics. The demand for precise positioning and error correction makes the image acquisition process expensive, complex calculations, etc., and it is difficult to deal with specific scenarios (such as white walls, etc.)

Images