IP Library Granted Patent US 12688712
Granted Patent B2
US 12688712 · App. 18/380,296 · Granted Jul 21, 2026

Seat belt detection

Inventors: Henrik Lind (Gothenburg, SE); Torsten Wilhelm (Gothenburg, SE); Ara Jafarzadeh (Gothenburg, SE); Joel Severin (Gothenburg, SE)
Assignee: Smart Eye AB
G06V20/59G06T7/13G06T7/73G06V40/10G06T2207/30196G06T2207/30268
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12688712
App. No.
18/380,296
Granted
Jul 21, 2026
Kind
B2
Abstract

A system and method for determining if an occupant of a vehicle has correctly applied a seat belt by acquiring an image of the occupant, the image including at least a shoulder portion of the occupant, identifying left and right shoulders and a longitudinal extension of a seat belt in the image, determining an intersection between the longitudinal extension and a line connecting the left and right shoulders, and, based on a relative position of the intersection and the shoulders, deducing whether the seat belt is correctly applied.

Claims (17)

1 . A computer implemented method for determining if an occupant of a vehicle has correctly applied a seat belt, the method comprising:

acquiring an image of the occupant, said image including at least a shoulder portion of the occupant;

identifying, in said image, positions of left and right shoulders;

identifying, in said image, a longitudinal extension of a seat belt;

determining an intersection between said longitudinal extension and a line connecting said left and right shoulders; and

based on a position of said intersection relative to the positions of the left and right shoulders, deducing whether said seat belt is correctly applied by comparing the intersection with the positions of the left and right shoulders and determining that the intersection is located between the left and right shoulders, wherein the step of deducing further includes determining that the intersection is located closer to one shoulder than a center point between said left and right shoulders.

2 . The method according to claim 1 , wherein the step of identifying left and right shoulders involves a human pose estimation algorithm, which is configured to identify locations of body joints.

3 . The method according to claim 1 , wherein the step of identifying a longitudinal extension of a seat belt includes edge detection in said image.

4 . A system for determining if an occupant of a vehicle has correctly applied a seat belt, the system comprising:

a camera for acquiring an image of the occupant, said image including at least a shoulder portion of the occupant;

a body analysis unit for identifying, in said image, positions of left and right shoulders;

a seat belt identifier configured to identify, in said image, a longitudinal extension of a seat belt; and

an assessment unit configured to:

determine an intersection between said longitudinal extension and a line connecting said left and right shoulders; and

based on a position of said intersection relative to the positions of said left and right shoulders, deduce whether said seat belt is correctly applied by comparing the intersection with the positions of the left and right shoulders and determining that the intersection is located between the left and right shoulders, wherein the assessment unit is further configured to determine that the intersection is located closer to one shoulder than a center point between said left and right shoulders.

5 . The system according to claim 4 , wherein the assessment unit is further configured to apply a human pose estimation algorithm adapted to identify locations of body joints in order to identify left and right shoulders.

6 . The system according to claim 4 , wherein the seat belt identifier is configured to identify said longitudinal extension using edge detection in said image.