SYSTEM AND METHOD FOR SCORING COLOR CANDIDATE POSES AGAINST A COLOR IMAGE IN A VISION SYSTEM

• US 20180225799A1
• Filed: 02/03/2017
• Published: 08/09/2018
• Est. Priority Date: 02/03/2017
• Status: Active Application

• Current Assignee: Cognex Corporation
• Original Assignee: Cognex Corporation
• Inventors: Davis, Jason, Michael, David J.
• Application Number:
  US15/424,767
• Publication Number:
  US 20180225799A1
• Time in Patent Office: Days
• CPC Class Codes: G06T 1/0014 Image feed-back for automatic industrial control, e.g. robot with camera robots B25J19/023
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 2207/10028 Range image; Depth image; 3D point clouds
• CPC Class Codes: G06T 2207/20016 Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
• CPC Class Codes: G06T 2207/20081 Training; Learning
• CPC Class Codes: G06T 2207/20092 Interactive image processing based on input by user
• CPC Class Codes: G06T 7/001 using an image reference approach
• CPC Class Codes: G06T 7/337 involving reference images or patches
• CPC Class Codes: G06T 7/60 Analysis of geometric attributes
• CPC Class Codes: G06T 7/73 using feature-based methods
• CPC Class Codes: G06T 7/74 involving reference images or patches
• CPC Class Codes: G06T 7/90 Determination of colour characteristics

Abstract

This invention provides a system and method for scoring a candidate pose in a geometric-pattern matching tool of a vision system by using trained color, grayscale and/or range (height) information (“color/grayscale/range”) in association with edge-aligned candidate poses. A trained pattern includes associated color/grayscale/range information in a set of test points. At runtime, a color, grayscale and/or range image is acquired and/or provided. A runtime pose is established with a coordinate space for the color/grayscale/range image with respect to the trained pattern, where the runtime pose is generated by an alignment tool. The color/grayscale/range test points are mapped onto the coordinate space for the image. The match is then determined at the respective mapped test points. Based on the test point match, a score is determined. The score is used in conjunction with the alignment result in runtime to accept or reject candidate poses from (e.g.) acquired images of runtime objects.

Claims

1. A method for scoring a candidate pose of a trained color pattern against a color image with a vision system having a camera assembly and a vision system processor, comprising the steps of:

providing, to the vision system processor, a trained color pattern with a set of color test points that represent color match information in the trained pattern, the color test points residing at least in regions of low gradient relative to an intensity image representation of the color image;

providing, to the vision system processor, a runtime color image of a scene;

establishing, with the vision system processor, a runtime pose with a coordinate space for the runtime color image with respect to the trained pattern where the runtime pose is generated by a geometric alignment process;

mapping, with the vision system processor, the color test points on the coordinate space for the image;

determining, with the vision system processor, the color match respectively at the mapped color test points; and

based on the color match, determining a color match score in at least a portion of the runtime color image.

2. The method as set forth in claim 1 wherein the step of determining the color match comprises determining a value for the color in a predetermined color space of the trained color pattern to the runtime color image at the mapped color test points.

3. The method as set forth in claim 1 wherein the step of establishing the pose comprises at least one of (a) employing alignment tools in the vision system to automatically align the runtime image to the trained color pattern and (b) obtaining information indicating the pose from a user input.

4. The method as set forth in claim 1, further comprising a mask that is applied to the runtime color image, wherein the mask indicates which areas of the runtime color image are evaluated for color match.

5. The method as set forth in claim 1, further comprising generating an intensity image from at least one of the trained color pattern and the runtime color image for use by the geometric alignment process.

6. The method as set forth in claim 1 wherein the color test points reside in regions of low gradient based upon a gradient magnitude threshold, and wherein the gradient magnitude threshold is established by at least one of (a) a user-input parameter and (b) a system-generated parameter.

7. A method for scoring a candidate pose of a trained grayscale pattern against a grayscale image with a vision system having a camera assembly and a vision system processor, comprising the steps of:

providing, to the vision system processor, a trained grayscale pattern with a set of grayscale test points that represent grayscale level match information in the trained pattern, the grayscale test points residing at least in regions of low gradient relative to an intensity image representation of the grayscale image;

providing, to the vision system processor, a runtime grayscale image of a scene;

establishing, with the vision system processor, a runtime pose with a coordinate space for the runtime grayscale image with respect to the trained pattern where the runtime pose is generated by a geometric alignment process;

mapping, with the vision system processor, the grayscale test points on the coordinate space for the image;

determining, with the vision system processor, the grayscale level match respectively at the mapped grayscale test points; and

based on the grayscale level match, determining a grayscale level match score in at least a portion of the runtime image.

8. The method as set forth in claim 7 wherein the step of establishing the pose comprises at least one of (a) employing alignment tools in the vision system to automatically align the runtime grayscale image to the trained grayscale pattern and (b) obtaining information indicating the pose from a user input.

9. The method as set forth in claim 7, further comprising a mask that is applied to the runtime greyscale image, wherein the mask indicates which areas of the runtime grayscale image are evaluated for grayscale level match.

10. The method as set forth in claim 7 wherein the grayscale test points reside in regions of low gradient based upon a gradient magnitude threshold, and wherein the gradient magnitude threshold is established by at least one of (a) a user-input parameter and (b) a system-generated parameter.

11. A method for scoring a candidate pose of a trained range image pattern against a range image with a vision system having a camera assembly and a vision system processor, comprising the steps of:

providing, to the vision system processor, a trained range image pattern with a set of range test points that represent range match information in the trained pattern, the range test points residing at least in regions of low gradient relative to an intensity image representation of the range image;

providing, to the vision system processor, a runtime range image of a scene;

establishing, with the vision system processor, a runtime pose with a coordinate space for the runtime range image with respect to the trained pattern where the runtime pose is generated by a geometric alignment process;

mapping, with the vision system processor, the range test points on the coordinate space for the image;

determining, with the vision system processor, the range match respectively at the mapped range test points; and

based on the range match, determining a range match score in at least a portion of the runtime image.

12. The method as set forth in claim 11 wherein the step of establishing the pose comprises at least one of (a) employing alignment tools in the vision system to automatically align the runtime range image to the trained range image pattern and (b) obtaining information indicating the pose from a user input.

13. The method as set forth in claim 11, further comprising a mask that is applied to the runtime range image, wherein the mask indicates which areas of the runtime range image are evaluated for range match.

14. The method as set forth in claim 11 wherein the range test points reside in regions of low gradient based upon a gradient magnitude threshold, and wherein the gradient magnitude threshold is established by at least one of (a) a user-input parameter and (b) a system-generated parameter.

15. The method as set forth in claim 11, further comprising generating an intensity image based upon at least one of the trained range image pattern and the runtime range image for use with the geometric alignment process.

16. A system for training a pattern with a vision system comprising:

a camera assembly and a vision system processor operatively connected to the camera assembly; and

a training module that receives training image data, identifies regions of low gradient magnitude and applies test points in the regions with respect to the image data based upon at least one of color, grayscale intensity and range up to a maximum test point count.

17. The system as set forth in claim 16 wherein the vision system processor includes a runtime alignment module arranged to (a) acquire runtime image data of a scene and (b) establish a runtime pose with a coordinate space for the runtime image data with respect to the trained pattern, where the runtime pose is generated by a geometric alignment process and (c) the color test points on the coordinate space for the runtime image data.

18. The system as set forth in claim 17 wherein the vision system processor includes a matching process that is arranged to (a) determine at least one of a color, grayscale intensity and range match between the trained pattern and the runtime image data and (b) determine a match score in at least a portion of the runtime image data.

19. The system as set forth in claim 18 further comprising an intensity image based upon at least one of the trained range image pattern and the runtime image data for use with the geometric alignment process.

20. The system as set forth in claim 18 wherein the pose is established based upon at least one of (a) alignment tools in the vision system that automatically align the runtime image to the trained pattern and (b) information indicating the pose from a user input.

21. The system as set forth in claim 18 further comprising a mask that is applied to the runtime image, wherein the mask indicates which areas of the runtime image are evaluated for matching.