Method and device for extraction of rod-shaped ground object

Abstract

The present invention relates to a method and device for extraction of a rod-shaped ground object. The method comprises: obtaining three-dimensional laser point cloud data, and removing low spatial points in the three-dimensional laser point cloud data to obtain a candidate point cloud set; dividing the candidate point cloud set to obtain an independent point cloud set corresponding to an independent ground object; identifying the rod-shaped ground object point cloud set in the independent point cloud set according to the projection of the independent point cloud set on a horizontal plane; and performing line extraction of the rod-shaped ground object point cloud set, and obtaining the upper and lower end points of the rod-shaped ground object according to the extracted elevation of the rod-shaped ground object point cloud on a straight line so as to effectively ensure the precision of the extracted rod-shaped ground object and satisfy the requirement of a high-precision map.

Description

Technical field [0001] The present invention relates to the field of computer technology, in particular to a method and device for extracting rod-shaped ground features. Background technique [0002] With the development of computer technology and automation technology, autopilot technology has received high attention. Autopilot technology cannot do without the support of high-precision maps. How to generate high-precision maps has become one of the bottlenecks of autonomous driving. High-precision rod-shaped feature data, such as street lamp poles, telephone poles, street sign poles and tree trunks, are also particularly important as an important part of high-precision maps. [0003] Existing rod-shaped feature extraction methods are often based on aerial photography or vehicle-mounted video extraction methods. Generally, they are first solved by aerial triangulation in photogrammetry to obtain the external orientation elements of each image, and then the rod-shaped features Mult...

AI Technical Summary

Problems solved by technology

[0003] Existing rod-shaped features extraction methods are often based on aerial photography or vehicle-mounted video extraction methods. Generally, the outer orientation elements of each image are obtained through aerial triangulation in photogrammetry, and then the rod-shaped features are extracted. Intersect multiple spaces in front to obtain their geometric coordinates. This method is limited by the resolution of the image, the calculation error of the internal and external orientation elements, and the extraction error of the feature points with the same name, etc., resulting in its absolute accuracy generally at the level of half a meter or even a few meters, which cannot meet high precision. Map accuracy requirements

Images