Artificial intelligence based registration support for environmental scans

    公开(公告)号:US11941793B2

    公开(公告)日:2024-03-26

    申请号:US17314408

    申请日:2021-05-07

    CPC classification number: G06T7/0002 G06T2207/10028 G06T2207/20081

    Abstract: A method for automatically determining quality of registration of landmarks includes training an artificial intelligence (AI) system to detect inaccurate registration of landmarks. Training the AI system uses training data that includes scans of an environment captured by a 3D measuring device from corresponding scan points. A first scan is registered with at least a second scan based on one or more landmarks captured in the first scan and the second scan. Further, a model is created to identify incorrect registration by analyzing the training data. The analysis detects a mismatch in a first instance of a landmark in the first scan and a second instance of said landmark in the second scan. The model is then used to evaluate registration of landmarks in live data, the live data including a set of scans, the result identifying accuracy level of the registration of landmarks.

    GLOBAL OPTIMIZATION METHODS FOR MOBILE COORDINATE SCANNERS

    公开(公告)号:US20240069203A1

    公开(公告)日:2024-02-29

    申请号:US18356871

    申请日:2023-07-21

    CPC classification number: G01S17/89

    Abstract: A mobile three-dimensional (3D) measuring system includes a 3D measuring device configured to capture 3D data in a multi-level architecture, and an orientation sensor configured to estimate an altitude. One or more processing units coupled with the 3D measuring device and the orientation sensor perform a method that includes receiving a first portion of the 3D data captured by the 3D measuring device. The method further includes determining a level index based on the altitude. The level index indicates a level of the multi-level architecture at which the first portion is captured. The level index is associated with the first portion. Further, a map of the multi-level architecture is generated using the first portion, the generating comprises registering the first portion with a second portion of the 3D data responsive to the level index of the first portion being equal to the level index of the second portion.

    CAPTURING ENVIRONMENTAL SCANS USING LANDMARKS BASED ON SEMANTIC FEATURES

    公开(公告)号:US20240004080A1

    公开(公告)日:2024-01-04

    申请号:US18469258

    申请日:2023-09-18

    CPC classification number: G01S17/894 G01S17/04 G01S7/4972

    Abstract: A method is provided that includes recording a landmark at a first scan position of a scanner, the landmark based at least in part on a semantic feature of scan data captured by the scanner. The semantic feature is identified using line-segments of the scan data. The method further includes capturing, by the scanner while moving through the environment, additional scan data at a second scan position. The method further includes, responsive to the scanner returning to the first scan position associated with the landmark, computing a measurement error. The method further includes correcting, using the measurement error, at least a portion of the scan data or the additional scan data.

    SYSTEM AND METHOD OF COMBINING THREE DIMENSIONAL DATA

    公开(公告)号:US20230252197A1

    公开(公告)日:2023-08-10

    申请号:US18106842

    申请日:2023-02-07

    Inventor: Göran Bergqvist

    CPC classification number: G06F30/13

    Abstract: According to one aspect of the disclosure a method for generating a three-dimensional model of an environment is provided. The method includes acquiring a first plurality of 3D coordinates of surfaces in the environment in a first coordinate frame of reference using a first measurement device, the first plurality of 3D coordinates including at least one subset of 3D coordinates of a target, the first measurement device optically measuring the plurality of 3D coordinates. A second plurality of 3D coordinates of the environment are acquired in a second frame of reference using a second measurement device, the second measurement device being operably disposed in a fixed relationship to the target. The second plurality of 3D coordinates with the first plurality of 3D coordinates are registered in the first coordinate frame of reference based at least in part on the at least one subset of 3D coordinates and the fixed relationship.

Patent Agency Ranking