Machine vision methods for identifying collinear sets of points from an image

• US 6,075,881 A
• Filed: 03/18/1997
• Issued: 06/13/2000
• Est. Priority Date: 03/18/1997
• Status: Expired due to Term

• Current Assignee: Cognex Corporation
• Original Assignee: Cognex Corporation
• Inventors: Loizeaux, Jane Alice, Foster, Nigel John
• Primary Examiner(s): MEHTA, BHAVESH M
• Application Number:
  US08/820,691
• Time in Patent Office: 1,183 Days
• Field of Search: 382/100, 382/151, 382/152, 382/153, 382/141, 382/171, 382/203, 382/154, 382/206, 382/276, 382/281, 382/286, 702/150, 702/151, 702/155, 702/156, 348/86
• US Class Current: 382/141
• CPC Class Codes: G06T 7/00 Image analysis
• CPC Class Codes: G06V 10/44 Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components

Abstract

The invention provides a two-phase method for identifying sets of collinear points in data points, e.g., derived from an image. During a first phase, the "angle-finding" phase, the angular orientation of the lines defined by the sets of points is estimated. During a second phase, the "line-finding" phase, sets of collinear points aligned at that angle are identified. The invention has application, for example, in finding parallel and perpendicular point sets in an image.

Claims

1. A machine vision method for identifying one or more sets of collinear points from data points, the method comprising the steps of:

A. searching among projections taken at a plurality of candidate angles determine an estimated angle of the sets of collinear points;

B. identifying the sets of collinear points from a projection of the data points at an angle based on the estimated angle; and

C. outputting a characteristic of a line formed by at least one of the sets of collinear points.

2. A method according to claim 1, wherein step (A) includes taking a projection of the data points at each of the candidate angles, and identifying one or more of the projections that produce highest density peaks.

3. A method according to claim 2, wherein step (A) includes the step of determining the estimated angle of the sets of collinear points from the candidate angles of plural projections that produce the highest density peaks.

4. A method according to claim 3, wherein step (A) includes the step of determining the estimated angle of the sets of collinear points by interpolating between the candidate angles of two projections that produce the two highest density peaks.

5. A method according to claim 2, wherein step (A) includes the step of filtering the projections to find those that produce the highest density peaks.

6. A method according to claim 5, wherein step (A) includes the step of using a filter sized in accord with an estimated width of a peak produced by a projection of the sets of collinear points at the candidate angle.

7. A method according to claim 1, wherein step (A) includes the step of defining the candidate angles in accord with any of (i) operator input, and (ii) an expected length and spacing of lines defined by the collinear sets of points.

8. A method according to claim 1, wherein step (A) includes the steps of

making a first determination of the estimated angle of the sets of collinear points by taking projections of the data points at a first set of candidate angles, that first set of candidate angles covering a first angular range at a first angular resolution, and

making successive determinations of the estimated angle of the sets of collinear points by taking projections of the data points at successive sets of candidate angles, each successive set of candidate angles covering a successively narrow angular range and being at a successively finer angular resolution than a prior set of candidate angles.

9. A method according to claim 1, wherein step (B) includes the steps of taking a projection of the data points at the estimated angle and finding one or more peaks in that projection corresponding to respective sets of collinear points.

10. A method according to claim 9, wherein step (B) includes the step of filtering the projection to find those peaks.

11. A method according to claim 10, wherein step (B) includes the step of using a filter sized in accord with an estimated width of a peak resulting from taking a projection of the set of collinear points at the estimated angle.

12. A method according to claim 9, wherein step (B) includes the step of merging into one set of collinear points two or more sets of collinear points that correspond to adjacent peaks in the projection.

13. A method according to claim 1, wherein step (C) includes the step of outputting, as the characteristic of a line formed by a set of collinear points, any of an angle of that line, a position of that line, and identities or locations of points making up that line.

14. A machine vision method for identifying one or more sets of collinear points from data points, the method comprising the steps of:

A. determining an estimated angle of the sets of collinear points by taking projections of the data points at a plurality of candidate angles, and by searching among the projection to find one or more projections that produce highest density peaks,

B. identifying one or more sets of collinear points by taking a projection of the data points at an angle based on the estimated angle, and finding one or more peaks in that projection corresponding to respective sets of collinear points, and

C. outputting a characteristic of a line formed by at least one of the sets of collinear points.

15. A machine vision method for identifying a set of collinear points from an data points, the method comprising the steps of:

A. making a first determination of an estimated angle of the sets of collinear points by taking projections of the data points at a first set of candidate angles covering a first angular range at a first angular resolution, and finding one or more projections producing highest density peaks,

B. making one or more successive determinations of the estimated angle by taking successive projections of the data points at successive sets of candidate angles, and finding one or more projections producing highest density peaks for each such set, each set of candidate angles covering a successively narrow angular range and being at a successively finer angular resolution than a prior set of candidate angles,

C. identifying one or more sets of collinear points by taking a projection of the data points at an angle based on the estimated angle, and finding one or more peaks in the projection corresponding to respective sets of collinear points, and

D. outputting a characteristic of a line formed by at least one of the sets of collinear points.

16. A method according to claim 15, wherein any of steps (A)-(C) include the step of filtering the projection to find one or more peaks therein.

17. A method according to claim 16, wherein including the step of using a filter sized in accord with an estimated width of a peak resulting from taking a projection of the corresponding set of collinear points.