Inspection method for inspecting an object and machine vision system

Patent metadata

Field	Value
Publication number	US-11978197-B2
Application number	US-202117512823-A
Country	US
Kind code	B2
Filing date	Oct 28, 2021
Priority date	Oct 28, 2021
Publication date	May 7, 2024
Grant date	May 7, 2024

Abstract

An inspection method includes comparing an inspection image of an inspection object to a reference image of a reference object, recognizing at least one inspection part in the inspection image and at least one reference part in the reference image, wherein the inspection part and the reference part correspond to each other, registering the inspection image onto the reference image using the inspection part and the reference part and providing a set of registration data, and checking for at least one error using the inspection image, the reference image, and the set of registration data.

First claim

What is claimed:

1. An inspection method, comprising:
comparing an inspection image of an inspection object to a reference image of a reference object;
recognizing at least one inspection part in the inspection image and at least one reference part in the reference image, wherein the inspection part and the reference part correspond to each other;
registering the inspection image onto the reference image using the inspection part and the reference part and providing a set of registration data, and
checking for at least one error using the inspection image, the reference image, and the set of registration data.

2. The inspection method according to claim 1 , wherein at least one of the inspection part or the reference part is recognized using a neural network.

3. The inspection method according to claim 1 , wherein the registering comprises:
performing a homography estimation based on at least one reference base point derived from the reference part and at least one inspection base point derived from the inspection part.

4. The inspection method according to claim 1 , wherein the at least one error comprises at least one of an incorrect part error, a part orientation error, an alignment error, a fixing element error, or a measurement error.

5. The inspection method according to claim 1 , wherein:
the inspection object is or corresponds to a composite construction object,
the composite construction object comprises a plurality of inspection construction parts,
the reference object is or corresponds to a reference composite construction object, and
the reference composite construction object comprises a plurality of reference construction parts.

6. The inspection method according to claim 5 , wherein the reference image comprises at least one of BIM data, CAD data, or a set of construction parts data.

7. The inspection method according to claim 5 , wherein checking for the at least one error comprises:
searching for presence of at least one characteristic property,
wherein the characteristic property includes a characteristic corresponding to a class of construction objects which the reference composite construction object belongs to.

8. The inspection method according to claim 7 , wherein the characteristic property is or at least comprises a horizontal or at least essentially horizontal construction part.

9. The inspection method according to claim 5 , further comprising:
computing a scaling factor for an element having known dimensions, and
searching for the element having the known dimensions in the inspection image based on the scaling factor.

10. The inspection method according to claim 9 , wherein the element having the known dimensions is or at least comprises at least one of a fiducial, a mark, or a tag.

11. The inspection method according to claim 1 , wherein checking for the at least one error comprises:
defining a focus region based on a reference part, and
comparing at least one inspection part being inside the focus region to the reference part.

12. The inspection method according to claim 11 , further comprising:
checking whether at least one inspection part inside the focus reason has at least one of an incorrect part error or a part orientation error.

13. The inspection method according to claim 11 , further comprising:
analyzing the focus region using a neural network.

14. The inspection method according to claim 11 , wherein checking for the at least one error comprises searching for at least one fixing element within the focus region.

15. The inspection method according to claim 14 , wherein checking for the at least one error comprises counting fixing elements within the focus region.

16. The inspection method according to claim 1 , further comprising:
presenting an overlaid image containing at least an area of the inspection image and at least an area of the reference image.

17. The inspection method according to claim 16 , wherein the overlaid image comprises at least one error-marking label.

18. The inspection method according to claim 17 , further comprising:
modifying a visibility of at least one of the areas of the inspection image or the reference image based on information received from a sliding button.

19. A machine vision system configured to inspect an inspection object comprising a plurality of inspection parts, the machine vision system comprising:
an inspection data interface configured to acquire an inspection image of the inspection object;
a reference data interface configured to acquire a reference image; and
at least one processor configured to compare an inspection image of the inspection object to a reference image of the reference object, the at least one processor comprising:
a part recognizer configured to recognize at least one inspection part in the inspection image and at least one reference part in the reference image,
registration logic configured to register the inspection image onto the reference image using the inspection part and the reference part and to provide a set of registration data, and
an error checker configured to check for at least one error using the inspection image, the reference image, and the set of registration data.

20. The machine vision system according to claim 19 , wherein the error checker is configured to check for at least one of an incorrect part error, a part orientation error, an alignment error, a fixing element error, or a measurement error.

21. The machine vision system according to claim 20 , wherein at least one of the part recognizer or the error checker comprises a neural network.

22. The machine vision system according to claim 21 , wherein each of the part recognizer and the error checker comprise a neural network.

23. The machine vision system according to claim 21 , wherein the neural network is configured to recognize a fixing element.

24. The machine vision system according to claim 21 , wherein the neural network is configured to recognize a support structure element.

25. The machine vision system according to claim 19 , further comprising:
a semantic foreground-filter configured to filter at least one of the inspection image or the reference image.

Assignees

• Hilti Ag

Inventors

• Sessinghaus Nadine
• Helmberger Michael
• Conway Alex

Classifications

• G06T7/001 Primary

  using an image reference approach · CPC title

• G06N3/04

  Architecture, e.g. interconnection topology · CPC title

• G06T7/0008

  checking presence/absence · CPC title

• G06T2207/30164

  Workpiece; Machine component · CPC title

• G06T2207/30204

  Marker · CPC title