Systems and methods for detecting an open door

Patent metadata

Field	Value
Publication number	US-2022227373-A1
Application number	US-202217658111-A
Country	US
Kind code	A1
Filing date	Apr 6, 2022
Priority date	Jun 24, 2020
Publication date	Jul 21, 2022
Grant date	—

Abstract

A computer-implemented method for navigating a host vehicle may include receiving an image frame acquired by an image capture device associated with the host vehicle; identifying in the image frame a representation of a target vehicle; determining an orientation indicator associated with the target vehicle; based on the determined orientation indicator, identifying a candidate region of the acquired image frame where a representation of a vehicle door of the target vehicle is expected in an open door condition; extracting the candidate region from the image frame; providing the candidate region to an open door detection network; determining a navigational action for the host vehicle in response to an indication from the open door detection network that the candidate region includes a representation of a door of the target vehicle in an open condition; and causing an actuator associated with the host vehicle to implement the navigational action.

First claim

What is claimed is:

1 . A system for navigating a host vehicle, the system comprising:
at least one processor comprising circuitry and a memory, wherein the memory includes instructions that when executed cause the at least one processor to:
receive an image frame acquired by an image capture device associated with the host vehicle, wherein the image frame is representative of an environment of the host vehicle;
identify in the image frame a representation of a target vehicle;
determine one or more orientation indicators associated with the target vehicle;
based on the determined one or more orientation indicators, identify at least one candidate region of the acquired image frame where a representation of a vehicle door of the target vehicle is expected in an open door condition;
extract the at least one candidate region from the image frame;
provide the extracted at least one candidate region to an open door detection network;
determine at least one navigational action for the host vehicle in response to an indication from the open door detection network that the extracted at least one candidate region includes a representation of a door of the target vehicle in an open condition; and
cause one or more actuators associated with the host vehicle to implement the at least one navigational action.

2 . The system of claim 1 , wherein the navigational action includes changing a steering direction for the host vehicle.

3 . The system of claim 1 , wherein the navigational action includes braking or slowing the host vehicle.

4 . The system of claim 1 , wherein the navigational action includes maintaining a current heading and speed of the host vehicle.

5 . The system of claim 4 , wherein the navigational action, including maintaining a current heading and speed of the host vehicle, is implemented if a predicted time to contact between the host vehicle and a detected open door of the target vehicle is greater than a threshold time.

6 . The system of claim 4 , wherein the navigational action, including maintaining a current heading and speed of the host vehicle, is implemented if a distance between the host vehicle and a detected open door of the target vehicle is greater than a threshold distance.

7 . The system of claim 1 , wherein the memory includes instructions that when executed cause the at least one processor to: compare a location of the detected open door of the target vehicle with a planned trajectory for the host vehicle and determine the at least one navigational action based on the comparison.

8 . The system of claim 1 , wherein the open door detection network is a trained neural network.

9 . The system of claim 1 , wherein the open door detection network is configured to output an indicator of a distance relative to a detected open door of the target vehicle.

10 . The system of claim 1 , wherein the one or more orientation indicators associated with the target vehicle indicate that a portion of a front side of the target vehicle is represented in the image frame and wherein the at least one candidate region of the image frame includes an image segment located to the right or to the left of a representation in the image frame of the front side of the target vehicle.

11 . The system of claim 10 , wherein a determination by the open door detection network of whether a door of the target vehicle is in an open condition is based, at least in part, on identification of exterior door features represented in the at least one candidate region of the image frame.

12 . The system of claim 1 , wherein the one or more orientation indicators associated with the target vehicle indicate that a portion of a rear side of the target vehicle is represented in the image frame and wherein the at least one candidate region of the image frame includes an image segment located to the right or to the left of a representation in the image frame of the rear side of the target vehicle.

13 . The system of claim 12 , wherein a determination by the open door detection network of whether a door of the target vehicle is in an open condition is based, at least in part, on identification of interior door features represented in the at least one candidate region of the image frame.

14 . The system of claim 1 , wherein the one or more orientation indicators associated with the target vehicle indicate that a portion of a side of the target vehicle is represented in the image frame and wherein the at least one candidate region of the image frame includes an image segment at least partially within an image envelope associated with a representation in the image frame of the side of the target vehicle.

15 . The system of claim 14 , wherein a determination by the open door detection network of whether a door of the target vehicle is in an open condition is based, at least in part, on identification of exterior door features. represented in the at least one candidate region of the image frame.

16 . The system of claim 1 , wherein the at least one candidate region is identified to the open door detection network by at least four coordinate points associated with positions of the image frame.

17 . The system of claim 1 , wherein the open door detection network is configured to output a classification of whether a door of the target vehicle is in an open condition.

18 . The system of claim 1 , wherein the open door detection network is configured to output a bounding box associated with a door of the target vehicle determined to be in an open condition.

19 . The system of claim 1 , wherein the open door detection network is configured to output one or more distance coordinates between at least a portion of the host vehicle. and a door of the target vehicle determined to be in an open condition.

20 . The system of claim 1 , wherein the open door detection network is configured to output a classification of whether a door of the target vehicle, at least partially occluded by a vehicle body, is in an open condition.

21 . The system of claim 1 , wherein the open door detection network is configured to output a classification of whether a door of the target vehicle, at least partially occluded by a pedestrian, is in an open condition.

22 . The system of claim 1 , wherein identification of the target vehicle is based, at least in part, on a determination of whether the target vehicle is moving.

23 . The system of claim 1 , wherein identification of the target vehicle is based, at least in part, on a determination of a lateral spacing between the host vehicle and the target vehicle.

24 . The system of claim 1 , wherein identification of the target vehicle is based, at least in part, on a determination of a longitudinal spacing between the host vehicle and the target vehicle.

25 . The system of claim 1 , wherein the open door detection network is configured to analyze the extracted at least one candidate region three or more times and output the indication that the extracted at least one candidate region includes a representation of a door of the target vehicle in an open condition after a determination that the three or more analyses all indicate an open door condition.

26 . A computer-implemented method for navigating a host vehicle, the computer-implemented method comprising:
receiving an image frame acquired by an image capture device associated with the host vehicle, wherein the image frame i

Assignees

• Mobileye Vision Technologies Ltd

Inventors

• Harouche Shahar
• Geller Idan
• Bentolila Jacob
• Saad Amir
• Lotan Roy

Classifications

• B60W2554/4029

  Pedestrians · CPC title

• B60W30/143

  Speed control (B60W30/16 takes precedence) · CPC title

• B60W60/0015

  specially adapted for safety · CPC title

• G01C21/3848

  Data obtained from both position sensors and additional sensors · CPC title

• B60W10/18

  including control of braking systems · CPC title