Point-cloud fusion

• US 9,562,971 B2
• Filed: 11/18/2013
• Issued: 02/07/2017
• Est. Priority Date: 11/22/2012
• Status: Expired due to Fees

• Current Assignee: Geosim Systems Ltd.
• Original Assignee: Geosim Systems Ltd.
• Inventors: Shenkar, Victor, Eilam, Yigal
• Primary Examiner(s): Kassa, Yosef
• Application Number:
  US14/646,394
• Publication Number:
  US 20150323672A1
• Time in Patent Office: 1,177 Days
• Field of Search: 345/419, 345/653
• US Class Current: 1/1
• CPC Class Codes: G01C 15/002 Active optical surveying means optical plumbing G01C15/105
• CPC Class Codes: G01S 17/86 Combinations of lidar systems with systems other than lidar, radar or sonar, e.g. with direction finders
• CPC Class Codes: G01S 17/87 Combinations of systems using electromagnetic waves other than radio waves
• CPC Class Codes: G01S 17/89 for mapping or imaging
• CPC Class Codes: G01S 7/4808 Evaluating distance, position or velocity data
• CPC Class Codes: G06T 17/05 Geographic models
• CPC Class Codes: G06T 2210/56 Particle system, point based geometry or rendering

Abstract

A method and a system for creating a collective point-cloud data from a plurality of local point-cloud data including the steps of: providing scanner data containing a plurality of local point-cloud data and relatively low-precision positioning data of the scanner providing the local point-cloud data, creating a relatively medium-precision collective point-cloud data from the plurality of local point-cloud data using external (medium-precision) positioning data associated with each of said plurality of local point-cloud data, and then creating a relatively high-precision collective point-cloud data from said medium-precision collective point-cloud data.

Claims

1. A method for creating a collective point-cloud data from a plurality of local point-cloud data, the method comprising the steps of:

providing scanner data comprising said plurality of local point-cloud data, wherein said local point-cloud data comprises relatively low-precision positioning data of said scanner;

creating a relatively medium-precision collective point-cloud data from said plurality of local point-cloud data; and

creating a relatively high-precision collective point-cloud data from said medium-precision collective point-cloud data;

wherein said step of creating a medium-precision collective point-cloud data comprises providing external (medium-precision) positioning data associated with each of said plurality of local point-cloud data.

2. The method of claim 1 wherein said low-precision positioning data comprises at least one of X, Y, Z coordinates, and yaw angle of a scanner providing said local point-cloud data;

and

wherein said step of creating a medium-precision collective point-cloud data comprises obtaining medium-precision data for at least one of said x, y, z coordinates, and yaw angle of said local point-cloud data.

3. The method of claim 2 wherein said external (medium-precision) positioning data comprises digital elevation model (DEM) and wherein said step of calculating medium-precision z coordinate for any of said local point-cloud data comprises using Z value from said DEM data.

4. The method of claim 2 wherein said step of calculating medium-precision for at least one of said x, y, z coordinates, and yaw angle comprises:

providing a global orthophoto data for said collective point-cloud data;

providing a local orthophoto data for each of said local point-cloud data;

aligning each of said local orthophoto data to said global orthophoto data to form a collectively-aligned local-orthophoto data; and

calculating said medium-precision values for at least one of said X, Y, and yaw for at least one of said local-orthophoto data based on said alignment of said local point-cloud data to said global orthophoto data.

5. The method of claim 4 wherein said step of providing said local orthophoto data comprises the steps of:

providing image data comprising of at least one of a ground image and a ceiling image, said image data associated with each of said local point-cloud data; and

converting said image data into a local orthophoto data.

6. The method of claim 2 wherein said step of aligning said local orthophoto data comprises the steps of:

providing inclination data to form roll and pitch values of said scanner providing said local point-cloud data for at least one of said local orthophoto data; and

using said roll and pitch values as medium-precision values of at least one of said local orthophoto data and said local point-cloud.

7. The method of claim 1 wherein said step creating a medium-precision collective point-cloud data comprises:

providing a global orthophoto data for said collective point-cloud data;

providing a local orthophoto data for said scanner when scanning said local point-cloud data;

aligning each of said local orthophoto data to said global orthophoto data; and

aligning said local point-cloud data, using said alignment of said local-orthophoto data to form said medium-precision collective point cloud data.

8. The method of claim 7 wherein said step of aligning said local orthophoto data comprises the steps of

obtaining a global digital elevation model (DEM) data associated with at least one of said global orthophoto and said collective point-cloud data;

and

obtaining z value from said DEM data for at least one of said global orthophoto data and said local-orthophoto data.

9. The method of claim 7 wherein said step of aligning said local orthophoto data comprises the steps of:

providing inclination data to form roll and pitch values of said scanner providing said local point-cloud data for at least one of said local orthophoto data; and

using said roll and pitch values as medium-precision values of at least one of said local orthophoto data and said local point-cloud.

10. The method of claim 1 wherein said step of creating a high-precision collective point-cloud data comprises:

extracting at least one feature from a first local point-cloud;

extracting at least one feature from a second local point-cloud;

matching at least one feature of said first local point-cloud with at least one feature of said second local point-cloud; and

applying a relative transformation comprising at least one of rotation and translation to at least one of said first and second local point-clouds to align said feature from said first local point-cloud with said feature from said second local point-cloud.

11. The method of claim 10 wherein at least one of said features is a planar patch.

12. The method of claim 10 wherein said step of matching said features comprises finding a group of features comprising at least one feature from said first local point-cloud and at least one feature from said second local point-clouds;

and

wherein said step of applying a relative transformation comprises finding a relative transformation that minimizes distances between said features from said first local point-cloud in said group and said features in said second local point-clouds in said group.

13. The method of claim 10 wherein said step of matching said features comprises features that are close according to said medium-precision collective point-cloud data.

14. The method of claim 10 wherein said step of matching said features comprises planar features that have a similar normal vector.

15. The method of claim 1 additionally comprising a step of:

geo-referencing said high-precision collective point-cloud data using relatively high-precision external positioning data.

16. The method of claim 15 said step of geo-referencing comprises:

obtaining said high-precision external positioning data for each of a plurality of said local point-cloud data.

17. The method of claim 15 wherein said high-precision external positioning data comprises global positioning system (GPS) data.

18. The method of claim 17 wherein said step of geo-referencing comprises:

selecting at least one GPS data according to a criterion selected from a group of criterions comprising;

GPS data comprising a predetermined minimal number of satellites;

GPS data comprising a predetermined minimal uninterrupted reception time of a minimal number of satellites; and

GPS data within a predetermined limit variance from an average of said GPS data.

19. A computer program product, stored on one or more non-transitory computer-readable medium, comprising instructions operative to cause a programmable processor of a network device to perform the steps of claim 1.

20. A method of creating a collective point-cloud data comprising a plurality of local point-clouds, said method comprising the steps of:

extracting at least one feature from a first local point-cloud;

extracting at least one feature from a second local point-cloud;

matching at least one feature of said first local point-cloud with at least one feature of said second local point-cloud; and

affecting a relative transformation comprising at least one of rotation and translation on at least one of said first and second local point-clouds to align said feature from said first local point-cloud with said feature from said second local point-cloud.

21. The method of claim 20 wherein at least one of said features is a planar patch.

22. The method of claim 20 wherein said step of matching said features comprises finding one or more groups of features, each comprising at least one feature from said first local point-cloud and at least one feature from said second local point-clouds;

and

wherein said step of affecting relative transformation comprises finding relative transformation that minimizes distances between said features from said first local point-cloud in each of said group and said features in said second local point-clouds in each of said group.

23. The method of claim 20 wherein said step of matching said features comprises features that are close according to a medium-precision collective point-cloud data.

24. The method of claim 20 wherein said step of matching said features comprises:

displaying said features to a user; and

enabling said user to affect at least one of selection and de-selection of features.

25. The method of claim 24 wherein said step of enabling said user to affect at least one of selection and de-selection of features comprises at least one of inclusion or removal of said features from said group.

26. The method of claim 20 wherein said step of matching said features comprises planar features that have a similar normal vector.

27. The methods of claim 20 wherein said step of matching said features comprises:

using a medium-precision collective point-cloud data,

setting parameters for determining distance threshold between cloud features,

displaying said features to a user,

matching features that are close according to said distance threshold, and

enabling said user to affect at least one of selection and de-selection of features.

28. The methods of claim 20 additionally comprising the step of:

setting parameters for determining planar vector similarity,

wherein said step of matching said features comprises determining planar vector similarity according to said parameter, and

enabling said user to affect at least one of selection and de-selection of features comprises least one of inclusion or removal of said features from said group based on said vector similarity.

29. A computer program product, stored on one or more non-transitory computer-readable medium, comprising instructions operative to cause a programmable processor of a network device to perform the steps of claim 20.