Method and apparatus for automatic discriminating and locating patterns such as finder patterns, or portions thereof, in machine-readable symbols

• US 6,097,839 A
• Filed: 03/10/1997
• Issued: 08/01/2000
• Est. Priority Date: 03/10/1997
• Status: Expired due to Fees

• Current Assignee: Intermec IP Corporation (Honeywell International Inc.)
• Original Assignee: Intermec IP Corporation (Honeywell International Inc.)
• Inventors: Liu, Lingnan
• Primary Examiner(s): Au, Amelia
• Assistant Examiner(s): Miller, Martin E
• Application Number:
  US08/814,001
• Time in Patent Office: 1,240 Days
• Field of Search: 235/462-464, 382/181-3, 382/197, 382/198, 382/267, 382/269
• US Class Current: 382/181
• CPC Class Codes: G06K 7/14 using light without selection of wavelength, e.g. sensing reflected white light G06K7/10831 - G06K7/1097 take precedence
• 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

A method and apparatus for locating geometric shapes or edges thereof in data collection symbols initially samples and stores an image of light reflected from the symbol. Thereafter, two routines are performed. A first routine performs low level vision processing by identifying linked points along edges, lines, curves or within shapes. A rapid pixel linking subroutine identifies edge pixels in symbol images lacking distortion. If the edge or shape suffers from distortions, then one or more distortion compensating subroutines locate sequential linked pixels despite such distortions. The resulting coordinates of linked points, which represent lines, curves and other geometric shapes, are then employed by a second routine which identifies patterns within the identified lines/curves. Based on these identified patterns, types of symbols from various symbologies can be identified and located within the stored image.

Claims

1. In a symbol reading apparatus capable of storing an image of a data collection symbol, the stored image being comprised of a plurality of pixels and the symbol having a predetermined pattern of geometric shapes, a method of locating portions of the shapes within the stored image, the method comprising:

locating a stalking pixel of a shape of the symbol within the stored image;

sequentially identifying adjacent pixels along an edge of the shape within the stored image from the starting pixel;

performing at least one distortion compensating subroutine selected from a plurality of distortion compensating subroutines if an adjacent pixel cannot be located, wherein at least one of the plurality of distortion compensating subroutines does not fill in the adjacent pixel;

selecting at least one linked pixel in the stored image based on the performing;

repeating the sequentially identifying, performing and selecting;

producing a set of pixels that identify at least a portion of the shape based on the repeating;

determining if the set of pixels contains representations of one of several selected patterns of geometric portions for corresponding data collection symbology types, a first selected pattern corresponding to the predetermined pattern of geometric shapes of the data collection symbol, wherein the selected patterns of geometric portions include lines and curves represented by a series of substantially adjacent pixels, and wherein the selected patterns do not represent widths of a plurality of shapes as measured along an intersecting sampling path; and

identifying the data collection symbol in the stored image if the detected geometric portions correspond to the first selected pattern.

2. The method of claim 1 wherein sequentially identifying includes:

determining a gradient direction from the starting pixel;

comparing the gradient direction to a predetermined value;

examining a plurality of pixels adjacent to the starting pixel, the adjacent pixels being in forward direction of the starting pixel based on the gradient direction;

selecting a forward pixel as a next edge starting pixel from the plurality of adjacent pixels; and

repeating the determining, comparing, examining and selecting a forward pixel until the starting pixel unfavorably compares to the predetermined value or a forward pixel cannot be selected.

3. The method of claim 1 wherein performing at least one distortion compensating subroutine includes:

determining a cross-sectional profile of the shape;

locating a center pixel of the cross-sectional profile;

selecting a pixel in a forward direction based on the center pixel;

determining a new cross-sectional profile of the shape at the pixel in the forward direction; and

repeating the locating, selecting and determining a new cross-sectional profile for a selected number of iterations.

4. The method of claim 1 wherein performing at least one distortion compensating subroutine includes:

determining an average intensity of pixels in a first area about the shape;

comparing each pixel in a second area to the average intensity, the second area being substantially smaller than the first area and being approximately centered about a selected pixel located under the sequentially identifying;

selecting a new adjacent pixel in the second area in a forward direction; and

repeating the comparing and selecting for a selected number of iterations.

5. The method of claim 1 wherein performing at least one distortion compensating subroutine includes:

adjusting an intensity of each pixel in an area based on a Gaussian smoothing filter, the area being approximately centered about a selected pixel located under the sequentially identifying;

examining the adjusted intensity of the pixels in the area;

selecting a new adjacent pixel in the area in a forward direction based on the step of examining; and

repeating the adjusting, examining and selecting for a selected number of iterations.

6. The method of claim 1 wherein sequentially identifying includes:

determining an edge direction of the shape;

identifying upper, upper right and right pixels adjacent to a current pixel as a forward three pixels if the edge direction is between 0 and 90 degrees;

identifying upper, tipper left and left pixels adjacent to a current pixel as a forward three pixels if the edge direction is between 90 and 180 degrees;

identifying left, lower left and lower pixels adjacent to a current pixel as a forward three pixels if the edge direction is between 180 and 270 degrees; and

identifying lower, lower right and right pixels adjacent to a current pixel as a forward three pixels if the edge direction is between 270 and 0 degrees.

7. The method of claim 1 wherein sequentially identifying includes:

identifying three pixels in a forward direction of a current pixel;

determining an edge magnitude based on an intensity of the three pixels; and

selecting a one of the three pixels having a greatest edge magnitude as a next adjacent pixel.

8. The method of claim 1 wherein determining if the set of pixels contains representations includes:

determining if the set of pixels contains at least two parallel lines, and if so, determining that the data collection symbol is a Code One symbol;

determining if the set of pixels contains at least two lines having ends intersecting at approximately 90 degrees, and if so, determining that the data collection symbol is a Data Matrix symbol; and

determining if the set of pixels contains at least a partially enclosed curve, and if so, determining that the data collection symbol is a MaxiCode symbol.

9. A computer implemented method of locating an image of a data collection symbol in a two-dimensional stored image, the two-dimensional stored image being represented by a plurality of points and the data collection symbol having a predetermined pattern of geometric shapes, the method comprising:

detecting geometric portions of the data collection symbol in the two-dimensional stored image based on groups points in the plurality of points;

determining, if the detected geometric portions correspond to one of several selected patterns of geometric portions for corresponding data collection symbology types, a first selected pattern corresponding, to the predetermined pattern of geometric shapes of the data collection symbol wherein the selected patterns of geometric portions includes lines and curves represented by a series of substantially adjacent pixels, and wherein the selected patterns do not represent widths of a plurality of shapes as measured along an intersecting sampling path; and

identifying the data collection symbol in the two-dimensional stored image if the detected geometric portions correspond to the first selected pattern.

10. The method of claim 9 wherein detecting geometric portions includes sequentially identifying adjacent points along an edge of the predetermined pattern within the stored image.

11. The method of claim 9 wherein sequentially identifying includes:

determining an edge direction from a current point;

comparing an associated value of the current point to a predetermined value;

examining a plurality of points adjacent to the current point;

selecting a new point as a new current point from the plurality of adjacent points; and

repeating the determining, comparing, examining and selecting a new point until the current point unfavorably compares to the predetermined value.

12. The method of claim 9 wherein detecting geometric portions includes performing at least one distortion compensating subroutine if a portion of one of the geometric shapes suffers from a distortion.

13. The method of claim 9 wherein detecting geometric portions includes:

determining a cross-sectional length of at least one of the geometric shapes;

locating a center point of the cross-sectional length;

selecting a new point based on the center point;

determining a new cross-sectional length of the geometric shape at the new point; and

repeating the locating, selecting and determining a new cross-sectional length for a selected number of iterations.

14. The method of claim 9 wherein detecting geometric portions includes:

determining a value corresponding to points in a first group of points about at least one of the geometric shapes;

comparing each point in a second group of points to the value, the second group being substantially smaller than the first group;

selecting a new point in the second group; and

repeating, at least once, the comparing and selecting.

15. The method of claim 9 wherein detecting geometric portions includes:

adjusting a value of each point in a group of points based on a smoothing function;

examining the adjusted value of the points in the group;

selecting a new point in the group based on the examining; and

repeating, at least once, the adjusting, examining and selecting.

16. The method of claim 9 wherein detecting geometric portions includes:

identifying three points from a current point;

determining an edge magnitude based on an intensity of the three points;

selecting a one of the three points having a greatest edge magnitude as a next adjacent point; and

repeating, at least once, the identifying three points, determining an edge and selecting.

17. The method of claim 9 wherein the data collection symbol is a Code One symbol, and wherein the step of determining includes the step of determining if the detected geometric portions contain the first selected pattern, wherein the first selected pattern includes at least two parallel lines.

18. The method of claim 9 wherein the data collection symbol is a Data Matrix symbol, and wherein the step of determining includes the step of determining if the detected geometric portions contain the first selected pattern, wherein the first selected pattern includes at least two lines intersecting at approximately 90 degrees.

19. The method of claim 9 wherein the data collection symbol is a MaxiCode symbol, and wherein determining includes determining if the detected geometric portions contain the first selected pattern, wherein the first selected pattern includes at least a partially enclosed curve.

20. In a symbol reading apparatus capable of storing an image of a data collection symbol, the stored image being comprised of a plurality of pixels and the symbol being comprised of a plurality of geometric shapes, a method of locating portions of the shapes within the stored image, the method comprising:

sequentially identifying adjacent pixels along an edge of the shape within the stored image;

performing at least one distortion compensating subroutine selected from a plurality of distortion compensating subroutines if an adjacent pixel cannot be located, wherein at least one of the plurality of distortion compensating subroutines does not fill in the adjacent pixel;

selecting at least one linked pixel in the stored image based on the performing;

producing a set of pixels that identify at least a portion of the shape based on the sequentially identifying and selecting; and

identifying a selected geometric shape from a set of shapes within the stored image based on the set of pixels, wherein the set of shapes includes lines and curves defined approximately by substantially adjacent pixels, and wherein the selected patterns do not represent widths of a plurality of shapes as measured along an intersecting sampling path.

21. The method of claim 20 wherein sequentially identifying includes:

determining an edge orientation from the starting pixel;

comparing a value of the starting pixel to a predetermined value;

examining a plurality of pixels adjacent to the starting pixel;

selecting a forward pixel as a next edge starting pixel from the plurality of adjacent pixels; and

repeating the determining, comparing, examining and selecting a forward pixel until the starting pixel unfavorably compares to the predetermined value.

22. The method of claim 20 wherein performing at least one distortion compensating subroutine includes:

determining a cross-section of the shape;

locating a center point of the cross-section;

selecting a pixel in a forward direction;

determining a new cross-section of the shape at the pixel in the forward direction; and

repeating the locating, selecting and determining a new cross-section for a selected number of iterations.

23. The method of claim 20 wherein performing at least one distortion compensating subroutine includes:

determining an average intensity of pixels in a first area about the shape;

comparing each pixel in a second area to the average intensity, the second area being substantially smaller than the first area and being approximately centered about a last adjacent pixel located under the sequentially identifying and before the performing;

selecting a new adjacent pixel in the second area in a forward direction; and

repeating the comparing and selecting for a selected number of iterations.

24. The method of claim 20 wherein performing at least on e distortion compensating subroutine includes:

adjusting an intensity of each pixel in an area based on a smoothing filter, the area being approximately centered about a last adjacent pixel located under the sequentially identifying and before the step of performing;

examining the adjusted intensity of the pixels in the area;

selecting a new adjacent pixel in the area in a forward direction based on the examining; and

repeating the adjusting, examining and selecting for a selected number of iterations.

25. The method of claim 20 wherein the sequentially identifying includes:

identifying three pixels in a forward direction of a current pixel;

determining an edge magnitude based on an intensity of the three pixels; and

selecting a one of the three pixels having a greatest edge magnitude as a next adjacent pixel.

26. In a symbol reading apparatus capable of storing an image of a data collection symbol, the stored image being comprised of a plurality of pixels and the symbol having a predetermined pattern of geometric shapes, a method of locating portions of the shapes within the stored image, the method comprising:

sequentially identifying adjacent pixels along an edge of the shape within the stored image from a starting pixel to produce a set of pixels that identify at least a portion of the shape; and

determining if the set of pixels contains representations of one of several selected patterns of geometric portions, which includes;

determining if the set of pixels contains at least two parallel lines, and if so, determining that the data collection symbol is a Code One symbol;

determining if the set of pixels contains at least two lines having ends intersecting at approximately 90°

, and if so, determining that the data collection symbol is a Data Matrix; and

determining if the set of pixels contains at least a partially enclosed curve, and if so, determining that the data collection symbol is a MaxiCode symbol.

27. A computer implemented method of locating an image of a data collection symbol in a two-dimensional stored image, the two-dimensional stored image being represented by a plurality of points and the data collection symbol having a pattern of geometric shapes, the method comprising:

detecting geometric portions of the data collection symbol in the two-dimensional stored image based on groups of points in the plurality of points, including;

determining a cross-sectional length of at least one of the geometric shapes;

locating a center point of the cross-sectional length;

selecting a new point based on the center point;

determining a new cross-sectional length of the geometric shape at the new point; and

repeating the locating, selecting and determining a new cross-sectional length, for at least one iteration; and

identifying the data collection symbol in the two-dimensional stored image.

28. A computer implemented method of locating an image of a data collection symbol in a two-dimensional stored image, the two-dimensional stored image being represented by a plurality of points and the data collection symbol having a predetermined pattern of geometric shapes, the method comprising:

detecting geometric portions of the data collection symbol in the two-dimensional stored image based on groups of points in the plurality of points;

determining if the detected geometric portions correspond to at least a partially enclosed curve that corresponds to one of several selected patterns of geometric portions; and

identifying the data collection symbol in the two-dimensional stored image as a MaxiCode symbol if the detected geometric portions include at least the partially enclosed curve.

29. In a symbol reading apparatus capable of storing an image of a data collection symbol, the stored image being comprised of a plurality of pixels and the symbol being comprised of a plurality of geometric shapes, a method of locating portions of the shapes within the stored image, the method comprising:

performing at least one distortion compensating subroutine selected from a plurality of distortion compensating subroutines if an adjacent pixel cannot be located, wherein at least one of the plurality of distortion compensating subroutines does not fill in the adjacent pixel, including;

determining a cross-section of at least one of the geometric shapes;

locating a center point of the cross-section;

selecting a pixel in a forward direction;

determining a new cross-section of the determined shape at the selected pixel in the forward direction; and

repeating the locating, selecting and determining a new cross-section for a selected number of iterations; and

identifying a selected geometric shape within the stored image.