Method and device of labeling laser point cloud
AssigneeBeijing Tusen Zhitu Technology Co., Ltd.
InventorsZhimeng Wang, Jieshu Li, He Zheng
Life of this patent
Estimating the term…
Abstract
The present application discloses a method and device of labeling laser point cloud. The method comprises: receiving data of a laser point cloud; constructing a 3D scene and establishing a 3D coordinate system corresponding to the 3D scene; converting a coordinate of each laser point in the laser point cloud into a 3D coordinate in the 3D coordinate system; mapping laser points included in the laser point cloud into the 3D scene respectively according to the 3D coordinate of the laser points; labeling the laser points in the 3D scene.
Claims20
A method of labeling a laser point cloud, comprising:
- receiving, by a device for labeling laser point clouds, data of the laser point cloud;
- constructing a 3D scene and establishing a 3D coordinate system corresponding to the 3D scene;
- converting a coordinate of each laser point in the laser point cloud into a 3D coordinate in the 3D coordinate system;
- mapping laser points in the laser point cloud into the 3D scene respectively according to using the 3D coordinates of the laser points;
- and
- labeling the laser points in the 3D scene, in which labeling the laser points in the 3D scene comprises:
- determining types to which the laser points in the 3D scene belong according to a first mode or a second mode:
- wherein the first mode comprises:
- generating a 3D selecting box according to first size information set by a user;
- moving a position of the 3D selecting box in the 3D scene according to a first moving instruction input by the user;
- and
- when receiving a labeling instruction carrying a target type, taking laser points currently falling within the 3D selecting box as target laser points and taking the target type in the labeling instruction as a type to which the target laser points belong;
- and
- wherein the second mode comprises:
- generating a ray by taking the position of a camera of at least one camera as a start point and taking a screen position point selected by the user as an end point;
- taking laser points meeting the following conditions as target laser points:
- a distance of a laser point from the start point is between a first distance threshold and a second distance threshold, and a vertical distance of the laser point from the ray is less than or equal to a third distance threshold;
- and
- when receiving a labeling instruction carrying a target type, taking the target type in the labeling instruction as a type to which the target laser points belong;
- and
- labeling the laser points by using labeling manners corresponding to the types to which the laser points belong, wherein different types correspond to different labeling manners.
A device for labeling laser point clouds, comprising:
- a receiving unit configured to receive data of a laser point cloud;
- a constructing unit configured to construct a 3D scene and establish a 3D coordinate system corresponding to the 3D scene;
- a converting unit configured to convert a coordinate of each laser point in the laser point cloud into a 3D coordinate in the 3D coordinate system;
- a mapping unit configured to map the laser points included in the laser point cloud into the 3D scene using the 3D coordinates of the laser points;
- a labeling unit configured to label the laser points in the 3D scene, the labeling unit comprising:
- a determining subunit configured to determine types to which the laser points in the 3D scene belong;
- and
- a labeling subunit configured to label the laser points by using labeling manners corresponding to the types to which the laser points belong, by:
- determining a labeling box pattern corresponding to the type to which the laser points belong according to a preset correspondence between types and labeling box patterns, including:
- selecting the labeling box pattern from a list of labeling box patterns with different colors according to the type to which the laser points belong;
- generating a 3D labeling box according to the selected labeling box pattern corresponding to the type to which the laser points belong;
- and
- labelling the laser points by using the 3D labeling box, wherein different types correspond to different labeling manners.
A device configured to label a laser point cloud, the device comprising a processor and a memory, wherein:
- the memory is configured to store at least one machine executable instruction, and the processor is configured to execute the at least one machine executable instruction to:
- receive data of a laser point cloud;
- construct a 3D scene and establish a 3D coordinate system corresponding to the 3D scene;
- convert a coordinate of each laser point in the laser point cloud into a 3D coordinate in the 3D coordinate system;
- map the laser points in the laser point cloud into the 3D scene using the 3D coordinates of the laser points;
- and
- label the laser points in the 3D scene, the instruction to label the laser points in the 3D scene comprising further instructions to:
- determine types to which the laser points in the 3D scene belongs;
- and
- label the laser points by using labeling manners corresponding to the types to which the laser points belong, wherein different types correspond to different labeling manners,
- wherein the processor executes the at least one machine executable instruction to label the laser points according to the labeling manner corresponding to the type to which the laser points belong, which comprises:
- determine a labeling box pattern corresponding to the type to which the laser points belong according to a preset correspondence between types to which the laser points belong and labeling box patterns, including:
- selecting the labeling box pattern from a list of labeling box patterns with different colors according to the type to which the laser points belong;
- generate a 3D labeling box according to the selected labeling box pattern corresponding to the type to which the laser points belong;
- and
- label the laser points by using the 3D labeling box.
Dependencies are inferred from the claim text.