# US 2025/0095286 A1 · A method and a system for generating 3d information relating to a scene or area comprising an object > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2025/0095286 A1 - Document ID: US20250095286A1 - Status: Published application - Kind: A1 - Application number: 18/468,056 - Filed: 2023-09-15 - Published: 2025-03-20 - Earliest claimed: 2023-09-15 - Assignees: Carnegie Robotics, LLC - Applicants: Carnegie Robotics, LLC - Inventors: Daniel David WILLIAMS; Stephen Paul Elizondo LANDERS; John Paul Thomas ATKINSON - Web page: https://patentgrep.com/patent/US20250095286A1 ## CPC - G06T 7/593: from stereo images - G06T 17/00: Three-dimensional [3D] modelling for computer graphics - G06T 7/13: Edge detection - G06T 7/50: Depth or shape recovery - G06T 2207/10012: Stereo images - G06T 2207/20084: Artificial neural networks [ANN] - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract Three-dimensional (3D) information is generated from two or more two-dimensional (2D) images. In the 2D images, one or more objects are identified, and either a corresponding portion of the 3D information representing the same object is identified and/or where the 3D information is corrected correspondingly. The correction may be the leaving out of 3D information representing the object or the correction of the 3D information based on knowledge of a shape of the object. ## Claims ### Claim 1 A method of generating object information from a stereo image pair, the method comprising: obtaining a stereo image pair comprising a first image and a second image of an area or scene comprising a plurality of objects, identifying, in the first image and/or the second image, a first object of the plurality of objects, generating, from the stereo image pair, 3D information relating to the scene or area, and identifying, based on the identified first object, a portion of the 3D information relating to the first object, the object information relating to the generated information. ### Claim 2 (depends on 1) A method according to claim 1, wherein: the identifying step comprises identifying also a second object of the plurality of objects, the generating step comprises excluding a portion of the stereo image pair representing the second object. ### Claim 3 (depends on 1) A method according to claim 1, further comprising the step of deriving shape information relating to a shape of a surface of the identified first object, wherein the generating step comprises generating, based on the shape information, a portion of the 3D information relating to the surface to have a shape corresponding to the shape of the surface. ### Claim 4 (depends on 1) A method according to claim 1, wherein: the identifying step comprises determining an edge or outline of at least a portion of the first object and deriving outline information, and the generating step comprises generating the 3D information based also on the outline information. ### Claim 5 (depends on 1) A method according to claim 1, the method further comprising the step of navigating in the area or scene based on the identified portion of the 3D information, wherein the first object is a stationary object. ### Claim 6 (depends on 2) A method according to claim 2, wherein the second object being dynamic, the method further comprising the step of navigating in the area or scene based on the generated 3D information. ### Claim 7 A system for generating object information from a stereo image pair, the system comprising: an obtaining element configured to obtain a stereo image pair comprising a first image and a second image of an area or scene comprising a plurality of objects, a controller configured to: identify, in the first image and/or the second image, a first object of the plurality of objects, generate, from the stereo image pair, 3D information relating to the scene or area, and identify, based on the identified first object, a portion of the 3D information relating to the first object, the object information relating to the generated information. ### Claim 8 (depends on 7) A system according to claim 7, wherein the controller is configured to: have the identifying step comprise identifying also a second object of the plurality of objects, have the generating step comprise excluding a portion of the stereo image pair representing the second object. ### Claim 9 (depends on 7) A system according to claim 7, wherein the controller is further configured to: derive shape information relating to a shape of a surface of the identified first object, and have the generating step comprise generating, based on the shape information, a portion of the 3D information relating to the surface to have a shape corresponding to the shape of the surface. ### Claim 10 (depends on 7) A system according to claim 7, wherein the controller is configured to have: the identifying step comprise determining an edge or outline of at least a portion of the first object and deriving outline information, and the generating step comprise generating the 3D information based also on the outline information. ### Claim 11 (depends on 7) A system according to claim 7, the controller further being configured to navigate in the area or scene based on the identified portion of the 3D information, wherein the first object is a stationary object. ### Claim 12 (depends on 8) A system according to claim 8, wherein the second object being dynamic, the controller being configured to navigate (a vessel, vehicle, robot . . . ) in the area or scene based on the generated 3D information. ### Claim 13 A method of generating 3D information from a stereo image pair, the method comprising: obtaining a stereo image pair comprising a first image and a second image of an area or scene comprising a plurality of objects, identifying, in at least one of the first image and the second image, a first object of the plurality of objects, generating, from the stereo image pair, the 3D information relating to the scene or area, and wherein the generating step excludes a portion of the stereo image pair representing the first object. ### Claim 14 A system for generating 3D information from a stereo image pair, the system comprising: an obtaining element configured to obtain a stereo image pair comprising a first image and a second image of an area or scene comprising a plurality of objects, a controller configured to: identify, in at least one of the first image and the second image, a first object of the plurality of objects, generate, from the stereo image pair, the 3D information relating to the scene or area, wherein the generation excludes a portion of the stereo image pair representing the first object. ### Claim 15 A method of generating 3D information from a stereo image pair, the method comprising: obtaining a stereo image pair comprising a first image and a second image of an area or scene comprising a plurality of objects, identifying, in the first image and/or the second image, a first object of the plurality of objects, obtaining information relating to the shape of the first object, generating, from the stereo image pair, 3D information relating to the scene or area, and adapting a portion of the 3D information representing the first object based on the obtained shape. ### Claim 16 A system for generating 3D information from a stereo image pair, the system comprising: an obtaining element configured to obtain a stereo image pair comprising a first image and a second image of an area or scene comprising a plurality of objects, a controller configured to: identify, in the first image and/or the second image, a first object of the plurality of objects, obtain information relating to the shape of the first object, generate, from the stereo image pair, 3D information relating to the scene or area, and adapt a portion of the 3D information representing the first object based on the obtained shape.