Method and device for fusing point cloud data and hybrid navigation system

Abstract

The invention discloses a method and device for fusing point cloud data. The method comprises the following steps: obtaining to-be-fused two point cloud data, wherein each of the two point cloud datacomprises a point three-dimensional coordinate and a point strength value, and both the to-be-fused two point cloud data contain point cloud data acquired in a same road area; respectively extractingall features from the two point cloud data according to the point strength values; determining all feature point pairs in the to-be-fused two point cloud data according to all the features extracted as well as point three-dimensional coordinate sets corresponding to all the features; fusing the two point cloud data according to all the feature point pairs determined. Through the method, the devicecan extract the features according to the strength values and calculate geometric centers of the features as feature points; the efficiency of the selecting the feature points is effectively improved; when each of the feature point pairs is selected, a selected position can be accurately matched.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to a method and device for point cloud data fusion. Background technique [0002] With the continuous progress and development of electronic map navigation technology, more and more research institutes and companies pay attention to vehicle automatic driving technology, and vehicle automatic driving technology must be based on high-precision maps. [0003] At present, in the process of making high-precision electronic maps, it is necessary to collect the location information of the road, and the location information of the road is usually collected by the laser carried by the collection vehicle to emit laser light to the surroundings, and receive the returned laser light when it hits an object, and record the laser return. The time and intensity of the laser return, and calculate the coordinate position of the object according to the time, determine what object the laser...

AI Technical Summary

Problems solved by technology

[0006] However, the inventor of the present application found that: when manually selecting feature points in a three-dimensional visual environment, it usually takes ten minutes to select each pair of feature points. It takes a lot of time and the efficiency is low. Moreover, since the three-dimensional visual environment cannot be infinitely enlarged, the selected positions of each pair of feature points manually selected cannot achieve a complete match, and there are errors.

Images