Granted Patent B2

US 12,496,966 · App. 18/473,208 · Granted Dec 16, 2025

Inventor: Mariusz Rychlik (Wegrzce Wielkie, PL)

Assignee: Aptiv Technologies AG

B60Q9/00B60R22/48G06T7/73G06V20/593G06V40/10G06V40/11G06V40/20G06V40/28B60R2022/4808B60R2022/4866B60W50/0098G06T2207/30196G06T2207/30268

Abstract

Claims (90)

1 . A perception-based monitoring method comprising:

receiving input data including a set of data points of an interior of a vehicle captured by a perception-based system mounted in the vehicle;

assigning, based on the set of data points, a set of predefined body keypoint positions (BKP) associated with a passenger, including wrist positions of the passenger and hand positions of the passenger, detected by the perception-based system;

computing a manipulation score indicative of an intention of the passenger to disengage a vehicle restraint based on the assigned set of predefined BKP in relation to the vehicle restraint, wherein the computing includes:

calculating a distance in 2D space of the set of data points between the vehicle restraint and at least one of the wrist positions or the hand positions; and

determining occlusion of the vehicle restraint, by at least one of the wrist positions or the hand positions, in 2D space of the set of data points;

in response to the manipulation score exceeding a predefined threshold, determining that the passenger possesses the intention to disengage the vehicle restraint; and

in response to determining that the passenger possesses the intention to disengage the vehicle restraint, at least one of:

adjusting at least one control function of a safety system of the vehicle; and

notifying at least one occupant of the vehicle of the intention to disengage the vehicle restraint.

2 . The method of claim 1 wherein the vehicle restraint includes at least one of a seatbelt and a door.

3 . The method of claim 2 wherein:

disengaging the seatbelt involves unbuckling a buckle of the seatbelt; and

disengaging the door involves opening the door using a handle of the door.

4 . The method of claim 1 wherein:

the vehicle restraint includes a seatbelt and a door;

disengaging the seatbelt involves unbuckling a buckle of the seatbelt; and

disengaging the door involves opening the door using a handle of the door.

5 . The method of claim 1 wherein:

the method includes, in response to determining that the passenger possesses the intention to disengage the vehicle restraint, notifying the at least one occupant of the vehicle; and

the notifying includes:

detecting existence of a set of predefined situations that require attention of a driver of the vehicle;

in response to detecting that at least one of the set of predefined situations exists:

preventing notification to the driver until existence of the at least one of the set of predefined situations is no longer detected, and

notifying the passenger; and

in response to detecting that none of the set of predefined situations exists, notifying the driver.

6 . The method of claim 1 wherein:

the assigning includes determining, from the set of data points, an eye gaze direction (EGD) of the passenger; and

the computing includes calculating an intersection of the EGD with the vehicle restraint in the set of data points.

7 . The method of claim 6 wherein:

the assigned set of predefined BKP and determined EGD are each given with a confidence value based on detection correctness factor associated with the perception-based system; and

the computing is weighted by the confidence values.

8 . The method of claim 1 wherein the input data includes at least one of:

status data of the safety system in operation to determine situations requiring driver's attention;

status data of the vehicle restraint of the passenger to prevent false positive notification;

vehicle instrumentation data of a window of the vehicle corresponding to the passenger, wherein the vehicle instrumentation data indicates whether the window is opening or closing to prevent false positive notification; and

calibration data on relative positions within the vehicle between the perception-based system and the vehicle restraint as parameters for computing the manipulation score.

9 . The method of claim 1 further comprising:

identifying false positives of the intention to disengage the vehicle restraint;

assigning a confidence coefficient to the identified false positives; and

amending the manipulation score based on the identified false positives weighted with the confidence coefficient.

10 . The method of claim 1 wherein:

the input data is generated based on a plurality of sets of data points acquired at different time frames over a predetermined time interval; and

the computing includes computing a respective manipulation score for each of the plurality of sets of data points.

11 . The method of claim 1 further comprising:

assigning, from the set of data points, a seat occupancy status (SOS) for each passenger seat; and

adapting the manipulation score based on the SOS.

12 . The method of claim 1 further comprising:

assigning, from the set of data points, a child seat presence (CSP) for each passenger seat; and

adapting the manipulation score based on the CSP.

13 . The method of claim 1 wherein the perception-based system includes at least one of:

a vision-based system mounted in an overhead area of the vehicle with a field of view covering one or more passenger seats; and

a Lidar system mounted in the overhead area with the field of view covering one or more passenger seats.

14 . The method of claim 13 wherein the perception-based system includes at least one of:

a second vision-based system located at a rear of the vehicle with a field of view of the one or more passenger seats from the rear of the vehicle; and

a second Lidar system located at the rear of the vehicle with a field of view of the one or more passenger seats from the rear of the vehicle.

15 . A non-transitory computer-readable medium comprising program instructions including:

receiving input data including a set of data points of an interior of a vehicle captured by a perception-based system mounted in the vehicle;

assigning, based on the set of data points, a set of predefined body keypoint positions (BKP) associated with a passenger, including wrist positions of the passenger and hand positions of the passenger, detected by the perception-based system;

computing a manipulation score indicative of an intention of the passenger to disengage a vehicle restraint based on the assigned set of predefined BKP in relation to the vehicle restraint, wherein the computing includes:

calculating a distance in 2D space of the set of data points between the vehicle restraint and at least one of the wrist positions or the hand positions; and

determining occlusion of the vehicle restraint by at least one of the wrist positions or the hand positions in 2D space of the set of data points;

in response to the manipulation score exceeding a predefined threshold, determining that the passenger possesses the intention to disengage the vehicle restraint; and

in response to determining that the passenger possesses the intention to disengage the vehicle restraint, at least one of:

adjusting at least one control function of a safety system of the vehicle; and

notifying at least one occupant of the vehicle of the intention to disengage the vehicle restraint.

16 . A perception-based system comprising a set of processors configured to execute instructions from memory, wherein the instructions include:

receiving input data including a set of data points of an interior of a vehicle captured by a perception-based system mounted in the vehicle;

assigning, based on the set of data points, a set of predefined body keypoint positions (BKP) associated with a passenger, including wrist positions of the passenger and hand positions of the passenger, detected by the perception-based system;

computing a manipulation score indicative of an intention of the passenger to disengage a vehicle restraint based on the assigned set of predefined BKP in relation to the vehicle restraint, wherein the computing includes:

calculating a distance in 2D space of the set of data points between the vehicle restraint and at least one of the wrist positions or the hand positions; and

determining occlusion of the vehicle restraint by at least one of the wrist positions or the hand positions in 2D space of the set of data points;

in response to the manipulation score exceeding a predefined threshold, determining that the passenger possesses the intention to disengage the vehicle restraint; and

in response to determining that the passenger possesses the intention to disengage the vehicle restraint, at least one of:

adjusting at least one control function of a safety system of the vehicle; and

notifying at least one occupant of the vehicle of the intention to disengage the vehicle restraint.

17 . A vehicle comprising the perception-based system according to claim 16 .

18 . A perception-based monitoring method comprising:

receiving input data including a set of data points of an interior of a vehicle captured by a perception-based system mounted in the vehicle;

assigning, based on the set of data points, a set of predefined body keypoint positions (BKP) associated with a passenger detected by the perception-based system;

computing a manipulation score indicative of an intention of the passenger to disengage a vehicle restraint based on the assigned set of predefined BKP in relation to the vehicle restraint;

in response to the manipulation score exceeding a predefined threshold, determining that the passenger possesses the intention to disengage the vehicle restraint; and

in response to determining that the passenger possesses the intention to disengage the vehicle restraint, at least one of:

adjusting at least one control function of a safety system of the vehicle; and

notifying at least one occupant of the vehicle of the intention to disengage the vehicle restraint,

wherein the input data includes at least one of:

status data of the safety system in operation to determine situations requiring driver's attention,

status data of the vehicle restraint of the passenger to prevent false positive notification,

vehicle instrumentation data of a window of the vehicle corresponding to the passenger, wherein the vehicle instrumentation data indicates whether the window is opening or closing to prevent false positive notification; and

calibration data on relative positions within the vehicle between the perception-based system and the vehicle restraint as parameters for computing the manipulation score.

Assignments (3)

Priority Claims (1)

Continuity (1)

References Cited (18)