Granted Patent B1

US 9,336,436 · App. 14/042,103 · Granted May 10, 2016

Abstract

Claims (56)

1. A method comprising:

receiving, at a computing device from a laser-based sensor, range data corresponding to an environment of a vehicle, wherein the range data comprises a plurality of data points in a point cloud that are indicative of respective positions of one or more objects in the environment relative to the vehicle;

detecting, by the computing device using the range data, one or more subsets of data points indicative of an upper-body region of a pedestrian from the plurality of data points in the point cloud, wherein the upper-body region comprises parameters corresponding to one or more of a head region and a chest region of the pedestrian;

identifying, using the one or more detected subsets of data points, a reflectivity ratio based on a first intensity level of data points corresponding to the head region and a second intensity level of data points corresponding to the chest region of the pedestrian;

determining whether the reflectivity ratio is in a range that indicates the upper-body region corresponds to a predefined reflectivity ratio range indicative of a pedestrian;

responsive to determining that the reflectivity ratio is in the predefined reflectivity ratio range, determining a position of the pedestrian relative to the vehicle;

based on the position of the pedestrian relative to the vehicle, determining a control strategy for the vehicle; and

causing, by the computing device, one or more systems of the vehicle to control the vehicle based on the control strategy.

2. The method of claim 1 , wherein detecting, by the computing device, the one or more subsets of data points indicative of the upper-body region of the pedestrian from the plurality of data points comprises:

determining whether or not the one or more subsets of data points correspond to a pedestrian in the environment based on a center of mass associated with a combination of the head region and the chest region.

3. The method of claim 1 , wherein detecting, by the computing device, the one or more subsets of data points indicative of the upper-body region of the pedestrian from the plurality of data points comprises:

determining a radial distance between the head region and the chest region of the pedestrian; and

based on the radial distance, determining whether or not the one or more subsets of data points correspond a pedestrian in the environment of the vehicle.

4. The method of claim 1 , wherein detecting, by the computing device, the one or more subsets of data points indicative of the upper-body region of the pedestrian from the plurality of data points comprises:

determining whether or not the one or more subsets of data points correspond to a pedestrian in the environment based on a measured point density of the head region and the chest region.

5. The method of claim 1 , further comprising:

determining an angle of maximum radial point based on a position of the chest region relative to a position of the head region; and

based on the angle of maximum radial point, determining an estimated orientation of the pedestrian.

6. The method of claim 5 , further comprising:

determining a potential intent of the pedestrian to change position based on the estimated orientation of the pedestrian.

7. The method of claim 1 , wherein detecting, by the computing device, the one or more subsets of data points indicative of the upper-body region of the pedestrian from the plurality of data points comprises:

based on a predefined pattern, determining the one or more subsets of data points from the plurality of data points that are indicative of the upper-body region of the pedestrian, wherein the predefined pattern is indicative of a shape and curvature associated with the head region and chest region.

8. An autonomous-vehicle system comprising:

one or more sensors;

a computer system configured to:

receive, from a laser-based sensor, range data corresponding to an environment of a vehicle, wherein the range data comprises a plurality of data points in a point cloud that are indicative of respective positions of one or more objects in the environment relative to the vehicle;

detect, using the range data, one or more subsets of data points indicative of an upper-body region of a pedestrian from the plurality of data points in the point cloud, wherein the upper-body region comprises parameters corresponding to one or more of a head region and a chest region of the pedestrian;

identify, using the one or more detected subsets of data points, a reflectivity ratio based on a first intensity level of data points corresponding to the head region and a second intensity level of data points corresponding to the chest region of the pedestrian;

determine whether the reflectivity ratio is in a range that indicates the upper-body region corresponds to a predefined reflectivity ratio range indicative of a pedestrian;

responsive to determining that the reflectivity ratio is in the predefined reflectivity ratio range, determine a position of the pedestrian relative to the vehicle;

based on the position of the pedestrian relative to the vehicle, determine a control strategy for the vehicle; and

cause one or more systems of the vehicle to control the vehicle based on the control strategy.

9. The system of claim 8 , wherein the function of detecting the one or more subsets of data points indicative of the upper-body region of the pedestrian from the plurality of data points comprises:

determining whether or not the one or more subsets of data points correspond to a pedestrian in the environment based on a center of mass associated with a combination of the head region and the chest region.

10. The system of claim 8 , wherein the function of detecting the one or more subsets of data points indicative of the upper-body region of the pedestrian from the plurality of data points comprises:

determining a radial distance between the head region and the chest region of the pedestrian; and

based on the radial distance, determining whether or not the one or more subsets of data points correspond a pedestrian in the environment of the vehicle.

11. The system of claim 8 , further comprising:

determining an angle of maximum radial point based on a position of the chest region relative to a position of the head region; and

based on the angle of maximum radial point, determining an estimated orientation of the pedestrian.

12. The system of claim 11 , further comprising:

determining a list of possible changing position actions of the pedestrian based on the estimated orientation of the pedestrian.

13. A non-transitory computer-readable medium having program instructions stored thereon, that when executed by at least one processor, cause the at least one processor to perform functions comprising:

receiving, from a laser-based sensor, range data corresponding to an environment of a vehicle, wherein the range data comprises a plurality of data points in a point cloud that are indicative of respective positions of one or more objects in the environment;

detecting, using the range data, one or more subsets of data points indicative of an upper-body region of a pedestrian from the plurality of data points in the point cloud, wherein the upper-body region comprises parameters corresponding to one or more of a head region and a chest region of the pedestrian;

identifying, using the one or more detected subsets of data points, a reflectivity ratio based on a first intensity level of data points corresponding to the head region and a second intensity level of data points corresponding to the chest region of the pedestrian;

determining whether the reflectivity ratio is in a range that indicates the upper-body region corresponds to a predefined reflectivity ratio range indicative of a pedestrian;

responsive to determining that the reflectivity ratio is in the predefined reflectivity ratio range, determining a position of the pedestrian relative to the vehicle;

based on the position of the pedestrian relative to the vehicle, determining a control strategy for the vehicle; and

causing one or more systems of the vehicle to control the vehicle based on the control strategy.

14. The non-transitory computer-readable medium of claim 13 , wherein detecting the one or more subsets of data points indicative of the upper-body region of the pedestrian from the plurality of data points comprises:

based on a predefined pattern, determining the one or more subsets of data points from the plurality of data points that are indicative of the upper-body region of the pedestrian, wherein the predefined pattern is indicative of a shape and curvature associated with the head region and chest region.

15. The non-transitory computer-readable medium of claim 13 , wherein the one or more subsets of data points comprise information based on a physical structure and curvature similar to a head region and a chest region of a pedestrian.

16. The non-transitory computer-readable medium of claim 13 , wherein detecting the one or more subsets of data points indicative of an upper-body region of the pedestrian from the plurality of data points comprises:

determining an angle of maximum radial point based on a position of the chest region relative to a position of the head region; and

based on the angle of maximum radial point, determining an estimated orientation of the pedestrian.

Assignments (6)