IP Library Granted Patent US 12,536,672
Granted Patent B2
US 12,536,672 · App. 18/039,807 · Granted Jan 27, 2026

Device and a method for representing the yaw orientation of a vehicle visible on an image

Inventor: Nikolay Chumerin (Brussels, BE)
Assignee: TOYOTA JIDOSHA KABUSHIKI KAISHA
G06T7/248G06T7/74G06T2207/10016G06T2207/30252
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 12,536,672
App. No.
18/039,807
Granted
Jan 27, 2026
Kind
B2
Abstract

A device and a method for representing the yaw orientation of a vehicle visible on an image, including: obtaining the position in the image of a rectangular bounding box surrounding the vehicle, obtaining the position of a vertical split line of the bounding box configured to be placed at a corner of the vehicle separating a first portion of the vehicle from a second portion of the vehicle, the first and second portion being either the front, the back, or a side of the vehicle, determining a signed horizontal shift of a horizontal position, associated with the vertical split line, with respect to a horizontal reference position, associated to the bounding box, normalizing the signed horizontal shift to obtain a value θ to be used in the representation of the pose of the vehicle.

Claims (31)

1 . A method for representing the yaw orientation of a vehicle visible on an image comprising:

obtaining the position in the image of a rectangular bounding box surrounding the vehicle,

obtaining the position of a vertical split line of the bounding box configured to be placed at a corner of the vehicle separating a first portion of the vehicle from a second portion of the vehicle, the first and second portions being either the front, the back, or a side of the vehicle, and at least one of the first and second portion is visible on the image, wherein the position associated with the vertical split line is the horizontal middle of either the first or the second portion,

determining a signed horizontal shift of a horizontal position, associated with the vertical split line, with respect to a horizontal reference position, associated to the bounding box, wherein the horizontal reference position is the horizontal middle of the bounding box or one lateral side of the bounding box; and

normalizing the signed horizontal shift to obtain a value θ to be used in the representation of the yaw orientation of the vehicle.

2 . The method of claim 1 , wherein the value θ to be used in the representation of the pose of the vehicle is calculated as:

θ= d/W

wherein d is the signed shift and W is a normalizing factor associated with the bounding box.

3 . The method of claim 2 , wherein W is the width of the bounding box.

4 . The method according to claim 1 , comprising determining a pose of the vehicle using the representation of the yaw orientation.

5 . The method of claim 1 , further comprising determining a yaw angle of the vehicle using the value θ.

6 . The method of claim 5 , wherein determining a yaw angle comprises using a given aspect ratio associated with the vehicle.

7 . The method of claim 6 , wherein determining the yaw angle comprises determining a function approximating the yaw angle taking as input the value θ and the aspect ratio associated with the vehicle, the function being:

a plane fitted on a set of given points associating a value of yaw angle with a value θ and an aspect ratio, or

constructed heuristically.

8 . A method for tracking the yaw orientation of a vehicle in a sequence of images comprising determining the value θ obtained by the method according to claim 1 for each image of the sequence of image.

9 . A device for representing the yaw orientation of a vehicle visible on an image comprising:

a module for obtaining the position in the image of a rectangular bounding box surrounding the vehicle,

a module for obtaining the position of a vertical split line of the bounding box configured to be placed at a corner of the vehicle separating a first portion of the vehicle from a second portion of the vehicle, the first and second portion being either the front, the back, or a side of the vehicle, wherein the position associated with the vertical split line is the horizontal middle of either the first or the second portion,

a module for determining a signed horizontal shift of a horizontal position, associated with the vertical split line, with respect to a horizontal reference position, associated to the bounding box, wherein the horizontal reference position is the horizontal middle of the bounding box or one lateral side of the bounding box; and

a module for normalizing the signed horizontal shift to obtain a value θ to be used in the representation of the yaw orientation of the vehicle.

10 . A vehicle comprising a device for representing the yaw orientation of a vehicle visible on an image the vehicle comprising at least one processor configured for:

obtaining the position in the image of a rectangular bounding box surrounding the vehicle,

obtaining the position of a vertical split line of the bounding box configured to be placed at a corner of the vehicle separating a first portion of the vehicle from a second portion of the vehicle, the first and second portion being either the front, the back, or a side of the vehicle, wherein the position associated with the vertical split line is the horizontal middle of either the first or the second portion,

determining a signed horizontal shift of a horizontal position, associated with the vertical split line, with respect to a horizontal reference position, associated to the bounding box, wherein the horizontal reference position is the horizontal middle of the bounding box or one lateral side of the bounding box; and

normalizing the signed horizontal shift to obtain a value θ to be used in the representation of the yaw orientation of the vehicle.

11 . A non-transitory computer-readable medium having recorded thereon a computer program including instructions that when executed by a processor cause the processor represent the yaw orientation of a vehicle visible on an image comprising:

obtaining the position in the image of a rectangular bounding box surrounding the vehicle,

obtaining the position of a vertical split line of the bounding box configured to be placed at a corner of the vehicle separating a first portion of the vehicle from a second portion of the vehicle, the first and second portions being either the front, the back, or a side of the vehicle, and at least one of the first and second portion is visible on the image, wherein the position associated with the vertical split line is the horizontal middle of either the first or the second portion,

determining a signed horizontal shift of a horizontal position, associated with the vertical split line, with respect to a horizontal reference position, associated to the bounding box, wherein the horizontal reference position is the horizontal middle of the bounding box or one lateral side of the bounding box; and

normalizing the signed horizontal shift to obtain a value θ to be used in the representation of the yaw orientation of the vehicle.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Aug 8, 2024
From: TOYOTA MOTOR EUROPE
To: TOYOTA JIDOSHA KABUSHIKI KAISHA
Reel/Frame 068226/0776 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 10, 2023
From: CHUMERIN, NIKOLAY
To: TOYOTA MOTOR EUROPE
Reel/Frame 064196/0027 →
Continuity (1)
Related Publication 20240005525A1 · Jan 4, 2024
References Cited (19)
US 12049170B2 · He · 2024 [cited by examiner]
US 20190335100A1 · Chen · 2019 [cited by examiner]
US 20200238991A1 · Aragon et al. · 2020 [cited by applicant]
CN 111081033A · 2020 [cited by examiner]
DE 102019002269A1 · 2020 [cited by examiner]
Mahendran et al.; “Monocular Object Orientation Estimation using Riemannian Regression and Classification Networks;” arXiv; 2018; pp. 1-24. [cited by applicant]
Kendall et al.; “PoseNet: a Convolutional Network for Real-Time 6-DOF Camera Relocalization;” Proceedings of the IEEE International Conference on Computer Vision; 2015; pp. 1-9. [cited by applicant]
Kendall et al.; “Modelling Uncertainty in Deep Learning for Camera Relocalization;” 2016 IEEE International Conference on Robotics and Automation (ICRA); IEEE; 2016; pp. 1-8. [cited by applicant]
Kendall et al.; “Geometric loss functions for camera pose regression with deep learning;” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition; 2017; pp. 1-10. [cited by applicant]
Mousavian et al.; “3D Bounding Box Estimation Using Deep Learning and Geometry;” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition; 2017; pp. 7074-7082. [cited by applicant]
Li et al.; “A Unified Framework for Multi-View Multi-Class Object Pose Estimation;” Proceedings of the European Conference on Computer Vision (ECCV); 2018; pp. 1-17. [cited by applicant]
Güler et al.; “DenseReg: Fully Convolutional Dense Shape Regression In-the-Wild;” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition; 2017; pp. 1-11. [cited by applicant]
Geiger et al.; Vision meets Robotics: The KITTI Dataset; The International Journal of Robotics Research 32.11 (2013); pp. 1231-1237. [cited by applicant]
L. Novak; “Vehicle detection and pose estimation for autonomous driving;” Diss. Master's thesis, Czech Technical University in Prague, 2017. [cited by applicant]
Huang et al.; “DenseBox: Unifying Landmark Localization with End to End Object Detection;” arXiv; 2015; pp. 1-13. [cited by applicant]
Xu et al.; “PointFusion: Deep Sensor Fusion for 3D Bounding Box Estimation;” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition; 2018; pp. 244-253. [cited by applicant]
Li et al.; “Vehicle Detection from 3D Lidar Using Fully Convolutional Network;” arXiv; 2016; pp. 1-8. [cited by applicant]
Chen et al.; “Multi-View 3D Object Detection Network for Autonomous Driving;” IEEE CVPR; 2017; pp. 1-9. [cited by applicant]
Aug. 18, 2021 Search Report issued in International Patent Application No. PCT/EP2020/084554. [cited by applicant]