IP Library Granted Patent US 12,573,064
Granted Patent B2
US 12,573,064 · App. 18/161,587 · Granted Mar 10, 2026

Monocular depth estimation parameter assistance and augmentation

Inventors: Emrah Akin Sisbot (Mountain View, CA); Seyhan Ucar (Mountain View, CA); Takamasa Higuchi (Mountain View, CA); Yongkang Liu (Mountain View, CA); Kentaro Oguchi (Mountain View, CA)
Assignees: TOYOTA MOTOR ENGINEERING & MANUFACTURING NORTH AMERICA, INC.; TOYOTA JIDOSHA KABUSHIKI KAISHA
G06T7/50G06V10/25G06V20/588G06T2207/30256
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 12,573,064
App. No.
18/161,587
Granted
Mar 10, 2026
Kind
B2
Abstract

Systems and methods are provided to determine optimal (image analysis) parameters for generating or adapting/adjusting a monocular depth estimation (MDE) depth map based on various factors, such as vehicle operating conditions, environmental conditions, road conditions, etc. After determining the optimal parameters to use, the optimal parameters can be stored on a cloud server or an edge server along with the vehicle location for use when another vehicle approaches the location to perform MDE without needing to independently determine the optimal parameters to use. Moreover, the use of such optimal image analysis parameters, enable a richer/enhanced view of the environment/area around or about a vehicle, where image analysis parameters may be used to more accurately interpret or otherwise characterize a depth map.

Claims (29)

1 . A method, comprising:

generating a depth map from an image of a scene;

identifying a portion of the scene corresponding to a road segment;

identifying a point of interest in the scene;

adjusting the depth map to de-emphasize the road segment;

determining an optimal image analysis parameter for the scene; and

transmitting, to a remote server, the optimal image analysis parameter and a geographic location at which the image was obtained.

2 . The method of claim 1 , further comprising extracting the road segment from the adjusted depth map.

3 . The method of claim 2 , wherein adjusting the depth map to de-emphasize the road segment is performed pursuant to extraction of the road segment from the depth map upon identification of the road segment.

4 . The method of claim 1 , further comprising prioritizing the identification of the point of interest relative to one or more other points of interest.

5 . The method of claim 1 , further comprising adjusting the depth map to emphasize the identified point of interest, wherein adjusting the depth map to emphasize the identified point of interest comprises highlighting the identified point of interest in a manner that visually emphasizes the identified point of interest to an operator of a vehicle, wherein the vehicle will encounter the identified point of interest while traversing the road segment.

6 . The method of claim 1 , wherein generating the depth map comprises generating the depth map using a monocular depth estimation (MDE) component based on the image of the scene received from a camera of a vehicle.

7 . The method of claim 6 , further comprising determining an operational parameter associated with the camera.

8 . The method of claim 7 , wherein determining the optimal image analysis parameter comprises generating a parameter value to offset at least one of the determined operational parameter and at least one aspect of the generated depth map.

9 . The method of claim 8 , wherein use of the determined optimal image analysis parameter increases accuracy of the generated depth map.

10 . A method, comprising:

receiving, at a first vehicle, an optimal image analysis parameter output by a remote server, the first vehicle traversing a road segment, the received optimal image analysis parameter having been generated pursuant to a second vehicle having previously traversed the road segment;

generating, at the first vehicle, a depth map from an image of a scene; and

adjusting, using the received optimal image analysis parameter, the depth map to de-emphasize the road segment.

11 . The method of claim 10 , wherein adjusting the depth map to de-emphasize the road segment is performed pursuant to extraction of the road segment from the depth map upon identification of the road segment.

12 . The method of claim 10 , further comprising prioritizing identification of a point of interest in the scene relative to one or more other points of interest.

13 . The method of claim 10 , further comprising adjusting the depth map to emphasize an identified point of interest in the scene, wherein adjusting the depth map to emphasize the identified point of interest comprises highlighting the identified point of interest in a manner that visually emphasizes the identified point of interest to an operator of the first vehicle, wherein the first vehicle will encounter the identified point of interest while traversing the road segment.

14 . The method of claim 10 , wherein generating the depth map comprises generating the depth map using a monocular depth estimation (MDE) component based on the image of the scene received from a camera of the first vehicle.

15 . The method of claim 14 , further comprising determining an operational parameter associated with the camera.

16 . The method of claim 15 , wherein the received optimal image analysis parameter comprises a parameter value to offset at least one of the determined operational parameter and at least one aspect of the generated depth map.

17 . The method of claim 16 , wherein use of the received optimal image analysis parameter increases accuracy of the generated depth map.

18 . The method of claim 1 , further comprising transmitting the determined optimal image analysis parameter to a remote server.

19 . The method of claim 18 , wherein the determined optimal image analysis parameter is transmitted to the remote server along with a location of a vehicle, the determined optimal image analysis parameter being associated with the location of the vehicle.

20 . The method of claim 18 , wherein transmitting the determined optimal image analysis parameter to the remote server is performed by a vehicle.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded May 28, 2026
From: TOYOTA JIDOSHA KABUSHIKI KAISHA
To: TOYOTA MOTOR ENGINEERING & MANUFACTURING NORTH AMERICA, INC.
Reel/Frame 074786/0005 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 30, 2023
From: SISBOT, EMRAH AKIN; UCAR, SEYHAN; HIGUCHI, TAKAMASA; LIU, YONGKANG; OGUCHI, KENTARO
To: TOYOTA MOTOR ENGINEERING & MANUFACTURING NORTH AMERICA, INC.; TOYOTA JIDOSHA KABUSHIKI KAISHA
Reel/Frame 062535/0409 →
Continuity (1)
Related Publication 20240257368A1 · Aug 1, 2024
References Cited (12)
US 1952010A · Gregory · 1934 [cited by examiner]
US 20150139533A1 · Wu · 2015 [cited by examiner]
US 20150304625A1 · Seto · 2015 [cited by applicant]
US 20190340729A1 · Haefner · 2019 [cited by examiner]
US 20200104612A1 · Chen · 2020 [cited by applicant]
US 20210112193A1 · Ter Haar · 2021 [cited by examiner]
US 20220148206A1 · Guizilini · 2022 [cited by applicant]
US 20220215577A1 · Kohata · 2022 [cited by examiner]
US 20220351225A1 · Girija · 2022 [cited by examiner]
US 20230281769A1 · Nanjegowd · 2023 [cited by examiner]
JP 2017142829A · 2017 [cited by applicant]
WO WO2022266131A1 · 2022 [cited by examiner]