Pedestrian path predictions

• US 9,786,177 B2
• Filed: 10/29/2015
• Issued: 10/10/2017
• Est. Priority Date: 04/10/2015
• Status: Active Grant

• Current Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Original Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Inventors: Ayvaci, Alper, Karasev, Vasiliy Igorevich, Heisele, Bernd, Zhu, Yingying
• Primary Examiner(s): Badii, Behrang
• Application Number:
  US14/926,030
• Publication Number:
  US 20160300485A1
• Time in Patent Office: 712 Days
• CPC Class Codes: B60G 9/02 the axle or housing being pivotally mounted on the vehicle , e.g. the pivotal axis being parallel to the longitudinal axis of the vehicle B60G9/003 takes precedence
• CPC Class Codes: G01S 3/786 the desired condition being maintained automatically
• CPC Class Codes: G01S 3/7864 T.V. type tracking systems
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06T 7/20 Analysis of motion motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51
• CPC Class Codes: G06T 7/277 involving stochastic approaches, e.g. using Kalman filters
• 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
• CPC Class Codes: G06V 40/103 Static body considered as a whole, e.g. static pedestrian or occupant recognition
• CPC Class Codes: G06V 40/23 Recognition of whole body movements, e.g. for sport training
• CPC Class Codes: G08G 1/01 Detecting movement of traffic to be counted or controlled G08G1/07 - G08G1/14 take precedence
• CPC Class Codes: G08G 1/0116 from roadside infrastructure, e.g. beacons
• CPC Class Codes: G08G 1/0133 for classifying traffic situation
• CPC Class Codes: G08G 1/09 Arrangements for giving variable traffic instructions railroad crossing signals B61L; reflectors E01F, G08B; indicating arrangements for variable information by selection or combination of individual elements G09F9/00
• CPC Class Codes: G08G 1/142 external to the vehicles
• CPC Class Codes: G08G 1/166 for active traffic, e.g. moving vehicles, pedestrians, bikes
• CPC Class Codes: G08G 9/02 Anti-collision systems

Abstract

Systems and techniques for pedestrian path predictions are disclosed herein. For example, an environment, features of the environment, and pedestrians within the environment may be identified. Models for the pedestrians may be generated based on features of the environment. A model may be indicative of goals of a corresponding pedestrian and predicted paths for the corresponding pedestrian. Pedestrian path predictions for the pedestrians may be determined based on corresponding predicted paths. A pedestrian path prediction may be indicative of a probability that the corresponding pedestrian will travel a corresponding predicted path. Pedestrian path predictions may be rendered for the predicted paths, such as using different colors or different display aspects, thereby enabling a driver of a vehicle to be presented with information indicative of where a pedestrian is likely to travel.

Claims

1. A system for pedestrian path predictions, comprising:

a sensor component identifying an environment, one or more features of the environment, and one or more pedestrians within the environment, wherein the sensor component observes one or more pedestrian states for the respective pedestrians;

a modeling component generating one or more models for one or more of the pedestrians based on one or more features of the environment, wherein a model of one or more of the models is indicative of one or more estimated goals of a corresponding pedestrian indicative of a desired destination for the corresponding pedestrian and one or more predicted paths for the corresponding pedestrian, wherein the model is generated based on one or more of the observed pedestrian states;

a prediction component determining two or more pedestrian path predictions for one or more of the pedestrians based on two or more corresponding predicted paths, wherein a pedestrian path prediction of one or more of the pedestrian path predictions is indicative of a probability that the corresponding pedestrian will travel a corresponding predicted path; and

an interface component rendering two or more of the pedestrian path predictions for two or more of the predicted paths, the rendering including a birds-eye-view of the environment, wherein each of the pedestrian path predictions is rendered to have a length associated with the probability that the corresponding pedestrian will travel on the corresponding predicted path, and wherein the interface component updates the rendering based on additional observed pedestrian states.

2. The system of claim 1, wherein the sensor component gathers one or more observations from one or more of the pedestrians.

3. The system of claim 2, wherein the modeling component infers one or more of the estimated goals for one or more of the models based on one or more of the observations.

4. The system of claim 1, wherein the modeling component generates one or more models for one or more of the pedestrians based on a Markov Decision Process (MDP).

5. The system of claim 1, wherein the modeling component generates one or more models for one or more of the pedestrians based on one or more pedestrian navigation preferences.

6. The system of claim 1, wherein the modeling component generates one or more models for one or more of the pedestrians based on one or more traffic rules.

7. The system of claim 1, wherein the modeling component generates one or more models for one or more of the pedestrians based on a status of a feature of one or more of the features of the environment.

8. The system of claim 1, comprising a score component generating one or more risk scores associated with one or more of the pedestrians indicative of a risk a vehicle equipped with the system for pedestrian path predictions poses an associated pedestrian or vice versa.

9. The system of claim 8, wherein the interface component renders one or more of the risk scores.

10. The system of claim 8, comprising a response component providing one or more operation actions based on one or more of the risk scores.

11. A method for pedestrian path predictions, comprising:

identifying an environment, one or more features of the environment, and one or more pedestrians within the environment;

observing one or more pedestrian states for the respective pedestrians;

generating one or more models for one or more of the pedestrians based on one or more features of the environment, wherein a model of one or more of the models is indicative of one or more estimated goals of a corresponding pedestrian indicative of a desired destination for the corresponding pedestrian and one or more predicted paths for the corresponding pedestrian, wherein the model is generated based on one or more of the observed pedestrian states;

determining two or more pedestrian path predictions for one or more of the pedestrians based on two or more corresponding predicted paths, wherein a pedestrian path prediction of one or more of the pedestrian path predictions is indicative of a probability that the corresponding pedestrian will travel a corresponding predicted path;

rendering two or more of the pedestrian path predictions for two or more of the predicted paths, the rendering including a birds-eye-view of the environment, wherein each of the pedestrian path predictions is rendered to have a length associated with the probability that the corresponding pedestrian will travel on the corresponding predicted path; and

updating the rendering based on additional observed pedestrian states.

12. The method of claim 11, comprising gathering one or more observations from one or more of the pedestrians.

13. The method of claim 12, comprising inferring one or more of the estimated goals for one or more of the models based on one or more of the observations.

14. The method of claim 11, comprising generating one or more models for one or more of the pedestrians based on a Markov Decision Process (MDP).

15. The method of claim 11, comprising generating one or more models for one or more of the pedestrians based on one or more pedestrian navigation preferences.

16. The method of claim 11, comprising generating one or more models for one or more of the pedestrians based on one or more traffic rules.

17. The method of claim 11, comprising generating one or more models for one or more of the pedestrians based on a status of a feature of one or more of the features of the environment.

18. The method of claim 11, comprising generating one or more risk scores associated with one or more of the pedestrians indicative of a risk a vehicle equipped with the system for pedestrian path predictions poses an associated pedestrian or vice versa.

19. The method of claim 18, comprising rendering one or more of the risk scores.

20. A system for pedestrian path predictions, comprising:

a sensor component identifying an environment, one or more features of the environment, and one or more pedestrians within the environment, wherein the sensor component observes one or more pedestrian states for the respective pedestrians;

a modeling component generating one or more models for one or more of the pedestrians based on a Markov Decision Process (MDP), wherein a model of one or more of the models is indicative of one or more estimated goals of a corresponding pedestrian indicative of a desired destination for the corresponding pedestrian and one or more predicted paths for the corresponding pedestrian, wherein the model is generated based on one or more of the observed pedestrian states;

a prediction component determining two or more pedestrian path predictions for one or more of the pedestrians based on two or more corresponding predicted paths, wherein a pedestrian path prediction of one or more of the pedestrian path predictions is indicative of a probability that the corresponding pedestrian will travel a corresponding predicted path; and

an interface component rendering two or more of the pedestrian path predictions for two or more of the predicted paths, the rendering including a birds-eye-view of the environment, wherein each of the pedestrian path predictions is rendered to have a length associated with the probability that the corresponding pedestrian will travel on the corresponding predicted path, and wherein the interface component updates the rendering based on additional observed pedestrian states.