How to distinguish whether a laser scan point belongs to an object in map or an unmodeled obstacle

asked 2021-02-13 02:17:24 -0500

brucexsk gravatar image

Hello to everyone. For example, in the picture below, the yellow points are laser scan data. The points in red circle belong to an object which was not modeled in map. And the other points are obstacles (such as wall, etc.) modeled in map. picture_url:

These points may not overlap with the objects in map strictly because of the positioning error or the map error. So I can't just use whether the point is within the black grid in map as the basis for classification. For example: [Points in the red rect are part of wall but they are not overlay with the black grids] picture_url:

So, how to classify the points, or distinguish them? Or, is there any algorithms can I use to solve this problem? Thanks !

edit retag flag offensive close merge delete