Back to results
US 12,578,472 B2Granted
Positioning data generation method, apparatus, and electronic device
AssigneeAlibaba Group Holding Limited
InventorsCheng Chen, Shuai Zhou, Qiang Xu
Life of this patent
Estimating the term…
Abstract
A positioning data generation method, an apparatus, and an electronic device are provided. The method comprises: obtaining laser point cloud data in a preset regional range on or by either side of the road; extracting laser point data of key points of a target object on or by either side of the road from the laser point cloud data, wherein the target object is a road object with a stable attribute on or by either side of the road; and storing the extracted laser point data of the key points of the target object as a piece of a plurality of pieces of positioning data of the road, the plurality of pieces of positioning data corresponding to a plurality of target objects on or by either side of the road.
Claims12
A positioning data generation method, comprising:
- obtaining laser point cloud data in a preset regional range on or by either side of a road;
- classifying the laser point cloud data as road-surface laser point cloud data and road-side laser point cloud data;
- fitting a road surface of the road according to the road-surface laser point cloud data, and
- adjusting, based on the fitted road surface, height values of laser points in the road-surface laser point cloud data or the road-side laser point cloud data to height values relative to the fitted road surface;
- extracting laser point data of key points of a target object on or by either side of the road from the laser point cloud data, wherein the target object is a road object on or by either side of the road, wherein the extracting comprises:
- extracting a key point of a ground marking on the road by:
- dividing the road-surface laser point cloud data into a plurality of grid cells according to a preset grid cell size;
- identifying, in each grid cell, a plurality of laser points corresponding to the ground marking based on reflectivity;
- and
- obtaining a key point of the ground marking in the grid cell based on the identified laser points;
- and
- extracting a key point of an upright object by a side of the road by:
- sorting laser points in the grid cell in ascending order of height values;
- if a height value difference of two adjacent laser points after sorting is greater than a difference threshold, deleting a laser point having a larger height value in the two adjacent laser points and one or more laser points following the laser point;
- in response to that (1) a smallest height value in the remaining laser points in the grid cell is smaller than a first height threshold and (2) a largest height value in the remaining laser points in the grid cell is greater than a second height threshold, obtaining the key point of the upright object in the grid cell based on the remaining laser points;
- storing the extracted laser point data of the key points of the target object as a piece of a plurality of pieces of positioning data of the road, the plurality of pieces of positioning data corresponding to a plurality of target objects on or by either side of the road.
13 more elements
A system for positioning, comprising a processor and a non-transitory computer-readable storage medium storing instructions executable by the processor to cause the system to perform operations comprising:
- obtaining laser point cloud data in a preset regional range on or by either side of a road;
- classifying the laser point cloud data as road-surface laser point cloud data and road-side laser point cloud data;
- fitting a road surface of the road according to the road-surface laser point cloud data, and
- adjusting, based on the fitted road surface, height values of laser points in the road-surface laser point cloud data or the road-side laser point cloud data to height values relative to the fitted road surface;
- extracting laser point data of key points of a target object on or by either side of the road from the laser point cloud data, wherein the target object is a road object on or by either side of the road, wherein the extracting comprises:
- extracting a key point of a ground marking on the road by:
- dividing the road-surface laser point cloud data into a plurality of grid cells according to a preset grid cell size;
- identifying, in each grid cell, a plurality of laser points corresponding to the ground marking based on reflectivity;
- and
- obtaining a key point of the ground marking in the grid cell based on the identified laser points;
- and
- extracting a key point of an upright object by a side of the road by:
- dividing the road-side laser point cloud data into a plurality of grid cells;
- in each grid cell:
- sorting laser points in the grid cell in ascending order of height values;
- if a height value difference of two adjacent laser points after sorting is greater than a difference threshold, deleting a laser point having a larger height value in the two adjacent laser points and one or more laser points following the laser point;
- in response to that (1) a smallest height value in the remaining laser points in the grid cell is smaller than a first height threshold and (2) a largest height value in the remaining laser points in the grid cell is greater than a second height threshold, obtaining the key point of the upright object in the grid cell based on the remaining laser points;
- storing the extracted laser point data of the key points of the target object as a piece of a plurality of pieces of positioning data of the road, the plurality of pieces of positioning data corresponding to a plurality of target objects on or by either side of the road.
15 more elements
A non-transitory computer-readable storage medium for positioning, configured with instructions executable by one or more processors to cause the one or more processors to perform operations comprising:
- obtaining laser point cloud data in a preset regional range on or by either side of a road;
- classifying the laser point cloud data as road-surface laser point cloud data and road-side laser point cloud data;
- fitting a road surface of the road according to the road-surface laser point cloud data, and
- adjusting, based on the fitted road surface, height values of laser points in the road-surface laser point cloud data or the road-side laser point cloud data to height values relative to the fitted road surface;
- extracting laser point data of key points of a target object on or by either side of the road from the laser point cloud data, wherein the target object is a road object on or by either side of the road, wherein the extracting comprises:
- extracting a key point of a ground marking on the road by:
- dividing the road-surface laser point cloud data into a plurality of grid cells according to a preset grid cell size;
- identifying, in each grid cell, a plurality of laser points corresponding to the ground marking based on reflectivity;
- and
- obtaining a key point of the ground marking in the grid cell based on the identified laser points;
- and
- extracting a key point of an upright object by a side of the road by:
- dividing the road-side laser point cloud data into a plurality of grid cells;
- in each grid cell:
- sorting laser points in the grid cell in ascending order of height values;
- if a height value difference of two adjacent laser points after sorting is greater than a difference threshold, deleting a laser point having a larger height value in the two adjacent laser points and one or more laser points following the laser point;
- in response to that (1) a smallest height value in the remaining laser points in the grid cell is smaller than a first height threshold and (2) a largest height value in the remaining laser points in the grid cell is greater than a second height threshold, obtaining the key point of the upright object in the grid cell based on the remaining laser points;
- storing the extracted laser point data of the key points of the target object as a piece of a plurality of pieces of positioning data of the road, the plurality of pieces of positioning data corresponding to a plurality of target objects on or by either side of the road.
15 more elements
Dependencies are inferred from the claim text.