Object bounding box estimation

• US 9,703,291 B1
• Filed: 01/04/2016
• Issued: 07/11/2017
• Est. Priority Date: 01/30/2012
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Waymo LLC (Alphabet Inc.)
• Inventors: Zhu, Jiajun, Haehnel, Dirk
• Primary Examiner(s): Tweel, Jr., John A
• Application Number:
  US14/986,813
• Time in Patent Office: 554 Days
• Field of Search: 3409951, 340933, 340937, 701 26, 701 1, 180169, 382103, 382104, 348148
• CPC Class Codes: B60K 31/0008 including means for detecting potential obstacles in vehicle path
• CPC Class Codes: B60W 2552/35 Road bumpiness, e.g. potholes
• CPC Class Codes: G01S 17/42 Simultaneous measurement of distance and other co-ordinates indirect measurement G01S17/46
• CPC Class Codes: G01S 17/931 of land vehicles
• CPC Class Codes: G05D 1/0088 characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours using knowledge based models G06N5/00
• CPC Class Codes: G05D 1/024 in combination with a laser lasers per se H01S
• CPC Class Codes: G05D 1/0274 using mapping information stored in a memory device navigation using map-matching G01C21/30
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/10028 Range image; Depth image; 3D point clouds
• CPC Class Codes: G06T 2207/30261 Obstacle
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/13 Edge detection
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle
• CPC Class Codes: G06V 20/58 Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads

Abstract

Aspects of the disclosure relate generally to maneuvering autonomous vehicles. Specifically, the vehicle may use a laser to collect scan data for a section of roadway. The vehicle may access a detailed map including the section of the roadway. A disturbance indicative of an object and including a set of data points data may be identified from the scan data based on the detailed map. The detailed map may also be used to estimate a heading of the disturbance. A bounding box for the disturbance may be estimated using the set of data points as well as the estimated heading. The parameters of the bounding box may then be adjusted in order to increase or maximize the average density of data points of the disturbance along the edges of the bounding box visible to the laser. This adjusted bounding box may then used to maneuver the vehicle.

Claims

1. A method for maneuvering a vehicle, the method comprising:

maneuvering, by one or more processors, the vehicle according to a control strategy;

selecting, by the one or more processors, a set of points from a plurality of data points generated by a sensor, each data point of the plurality of data points representing a three-dimensional location of a point corresponding to a surface of an object located on a section of roadway, the set of points being indicative of an object in the section of roadway;

estimating, by the one or more processors, a bounding box for the set of points based on a detailed map of the section of roadway;

adjusting, by the one or more processors, a parameter of the estimated bounding box based on the three-dimensional locations of at least some of the set of points;

adjusting, by the one or more processors, the control strategy based on the adjusted parameter; and

maneuvering, by the one or more processors, the vehicle according to the adjusted control strategy.

2. The method of claim 1, wherein the set of data points is further selected by selecting data points of the plurality of data points that satisfy a threshold elevation above a surface of the section of roadway identified in the detailed map.

3. The method of claim 1, wherein the set of data points is further selected by selecting data points of the plurality of data points that do not correspond to objects identified in the detailed map.

4. The method of claim 1, wherein the control strategy include slowing the vehicle down in a first way and the adjusted control strategy includes slowing the vehicle down in a second way different from the first way.

5. The method of claim 4, wherein the second way includes slowing the vehicle down more gradually than the first way.

6. The method of claim 1, wherein the control strategy includes maintaining at least a first distance from the object and the adjusted control strategy includes maintaining at least a second distance from the object, the first distance being different from the first distance.

7. The method of claim 6, wherein the first distance is greater than the second distance.

8. The method of claim 1, wherein adjusting the parameter includes:

projecting the set of data points onto a two-dimensional intensity map within the estimated bounding box; and

identifying an edge of the bounding box visible to the sensor, and

wherein the parameter is adjusted in order to increase a number of data points of the set of data points projected along the identified edge.

9. The method of claim 1, further comprising:

receiving additional data points generated by the sensor corresponding to the object;

using the estimated bounding box and adjusted parameter to estimate a new bounding box for the object based on the received additional data points; and

using the new bounding box to maneuver the vehicle.

10. The method of claim 1, further comprising:

receiving sensor data from a second sensor of a different type from the sensor, and

wherein the received sensor data from the second sensor is further used to estimate the estimated bounding box.

11. A system for maneuvering a vehicle, the system comprising one or more computing devices having one or more processors, the one or more computing devices being configured to:

maneuver the vehicle according to a control strategy;

select a set of points from a plurality of data points generated by a sensor, each data point of the plurality of data points representing a three-dimensional location of a point corresponding to a surface of an object located on a section of roadway, the set of points being indicative of an object in the section of roadway;

estimate a bounding box for the set of points based on a detailed map of the section of roadway;

adjust a parameter of the estimated bounding box based on the three-dimensional locations of at least some of the set of points;

adjust the control strategy based on the adjusted parameter; and

maneuver the vehicle according to the adjusted control strategy.

12. The system of claim 11, wherein the one or more computing devices are further configured to select the set of data points further by selecting data points of the plurality of data points that satisfy a threshold elevation above a surface of the section of roadway identified in the detailed map.

13. The system of claim 11, wherein the one or more computing devices are further configured to select the set of data points further by selecting data points of the plurality of data points that do not correspond to objects identified in the detailed map.

14. The system of claim 11, wherein the control strategy include slowing the vehicle down in a first way and the adjusted control strategy includes slowing the vehicle down in a second way different from the first way.

15. The system of claim 14, wherein the second way includes slowing the vehicle down more gradually than the first way.

16. The system of claim 11, wherein the control strategy includes maintaining at least a first distance from the object and the adjusted control strategy includes maintaining at least a second distance from the object, the first distance being different from the first distance.

17. The system of claim 16, wherein the first distance is greater than the second distance.

18. The system of claim 11, further comprising the sensor.

19. The system of claim 11, further comprising the vehicle.

20. A tangible computer-readable storage medium on which computer readable instructions of a program are stored, the instructions, when executed by one or more processors, cause the one or more processors to perform a method of maneuvering a vehicle, the method comprising:

maneuvering the vehicle according to a control strategy;

selecting a set of points from a plurality of data points generated by a sensor, each data point of the plurality of data points representing a three-dimensional location of a point corresponding to a surface of an object located on a section of roadway, the set of points being indicative of an object in the section of roadway;

estimating a bounding box for the set of points based on a detailed map of the section of roadway;

adjusting a parameter of the estimated bounding box based on the three-dimensional locations of at least some of the set of points;

adjusting the control strategy based on the adjusted parameter; and

maneuvering the vehicle according to the adjusted control strategy.