Lane line creation for high definition maps for autonomous vehicles
An HD map system represents landmarks on a high definition map for autonomous vehicle navigation, including describing spatial location of lanes of a road and semantic information about each lane, and along with traffic signs and landmarks. The system generates lane lines designating lanes of roads based on, for example, mapping of camera image pixels with high probability of being on lane lines into a three-dimensional space, and locating/connecting center lines of the lane lines. The system builds a large connected network of lane elements and their connections as a lane element graph. The system also represents traffic signs based on camera images and detection and ranging sensor depth maps. These landmarks are used in building a high definition map that allows autonomous vehicles to safely navigate through their environments.
1 . A method comprising:
identifying, based at least on sensor data, a plurality of clusters of points in which individual clusters of the plurality of clusters respectively correspond to individual lane line segments of a plurality of lane line segments corresponding to a lane line;
identifying a plurality of center lines of the plurality of clusters in which individual center lines of the plurality of center lines are respectively identified as being respective lines running substantially along lengths of and through centers of the individual clusters of the plurality of clusters;
identifying the lane line based at least on the plurality of center lines; and
causing performance of one or more control operations corresponding to a machine based at least on the lane line as identified.
2 . The method of claim 1 , wherein the sensor data includes image data.
3 . The method of claim 1 , wherein the identifying of one or more clusters of points of the plurality of clusters of points is based at least on relative proximities of points with respect to one another.
4 . The method of claim 1 , wherein one or more clusters of points of the plurality of clusters of points are identified from a set of points classified as generally corresponding to lane lines.
5 . The method of claim 1 , wherein the lane line is identified by at least connecting endpoints corresponding to the plurality of center lines.
6 . The method of claim 1 , wherein one or more center lines of the plurality of center lines are identified based at least on one or more respective geometric fit lines determined with respect to one or more corresponding clusters of points.
7 . The method of claim 1 , further comprising generating, based at least on the lane line as identified, a lane element graph that includes lane elements and connections between lane elements.
8 . A system comprising:
one or more processors to perform operations comprising:
identifying a plurality of center lines in which individual center lines of the plurality of center lines are respectively identified as being respective lines running lengthwise substantially through respective centers of individual clusters of a plurality of clusters of points in which individual clusters of the plurality of clusters respectively correspond to individual lane line segments of a plurality of lane line segments corresponding to a lane line;
identifying the lane line based at least on the plurality of center lines;
generating, based at least on the lane line as identified, a lane element graph that includes one or more lane elements of a lane corresponding to the lane line; and
causing performance of one or more control operations corresponding to a machine based at least on the lane element graph.
9 . The system of claim 8 , wherein the generating of the lane element graph includes identifying the lane line as a boundary of at least one lane element of the one or more lane elements.
10 . The system of claim 8 , wherein the lane element graph includes at least one of:
one or more semantic associations between the one or more lane elements; or
one or more features respectively corresponding to the one or more lane elements.
11 . The system of claim 8 , wherein one or more clusters of points of the plurality of clusters of points are identified based at least on relative proximities of points with respect to one another.
12 . The system of claim 8 , wherein one or more clusters of points of the plurality of clusters of points are identified from a set of points classified as generally corresponding to lane lines.
13 . The system of claim 8 , wherein one or more center lines of the plurality of center lines are identified based at least on one or more respective geometric fit lines determined with respect to one or more corresponding clusters of points.
14 . One or more processors comprising:
processing circuitry to perform operations comprising:
identifying one or more clusters of points respectively corresponding to individual lane line segments of one or more lane line segments corresponding to a lane line;
identifying the lane line based at least on one or more center lines that respectively run substantially through one or more respective centers of the one or more clusters of points;
generating, based at least on the lane line as identified, a lane element graph that includes lane elements and connections between lane elements; and
causing performance of one or more control operations corresponding to a machine based at least on the lane element graph.
15 . The one or more processors of claim 14 , wherein the generating of the lane element graph includes identifying the lane line as a boundary of at least one lane element of the one or more lane elements.
16 . The one or more processors of claim 14 , wherein the identifying of at least one cluster of points is based at least on relative proximities of points with respect to one another.
17 . The one or more processors of claim 14 , wherein at least one cluster of points is identified from a set of points classified as generally corresponding to lane lines.
18 . The one or more processors of claim 14 , wherein the lane line is identified by at least connecting endpoints corresponding to the one or more center lines.
19 . The one or more processors of claim 14 , wherein at least one center line of the one or more center lines is identified based at least on one or more respective geometric fit lines determined with respect to a corresponding cluster of points.
20 . The one or more processors of claim 19 , wherein the one or more center lines are identified based at least on removal of one or more outlier points corresponding to the one or more geometric fit lines.