Method and apparatus for high speed identification of objects having an identifying feature

• US 5,742,037 A
• Filed: 03/07/1996
• Issued: 04/21/1998
• Est. Priority Date: 03/07/1996
• Status: Expired due to Term

• Current Assignee: Cognex Corporation
• Original Assignee: Cognex Corporation
• Inventors: Li, David Y., Scola, Joseph R.
• Primary Examiner(s): Hajec, Donald T.
• Assistant Examiner(s): FRECH, KARL D
• Application Number:
  US08/612,032
• Time in Patent Office: 775 Days
• Field of Search: 235/454, 235/462, 235/463
• US Class Current: 235/454
• 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: G06K 7/1452 detecting bar code edges

Abstract

A method and apparatus for identifying an object having an identifying feature includes a two step process for detecting the identifying feature. The process involves locating the feature based upon a first characteristic of the feature and then verifying that feature is the identifying feature based upon a second characteristic of the feature. Preferably, the first step of the process is less discriminating and substantially faster than the second step. This permits the process to be performed rapidly without reducing accuracy.

Claims

1. A method for identifying an object having an identifying feature, said identifying feature having at least one predefined characteristic, said method comprising the steps of:

1) creating a representation of said object;

2) locating at least one feature in said representation of said object having a corresponding characteristic similar to said at least one characteristic of said identifying feature and creating a second representation of at least one other characteristic of said located feature, the step of locating including the substeps of;

sampling at least one portion of said representation of said object at a predefined position in said representation, including sampling, including sampling at predetermined locations across said representation of said object in a predefined direction, said locations being spaced apart by a distance corresponding to at least one characteristic of said identifying feature; and

creating said second representation of said sampled portion of said representation consisting of a representation of at least one other characteristic of said located feature;

3) normalizing at least a portion of said second representation of said at least one other characteristic of said located feature, to create a normalized representation of said feature;

4) comparing said normalized representation of said at least one other characteristic of said feature to a predefined representation of a corresponding characteristic of said identifying feature.

2. A method according to claim 1, wherein said distance is less than a dimension of said identifying feature in said predefined direction.

3. A method according to claim 2, wherein said distance is less than one half a dimension of said identifying feature in said predefined direction.

4. A method according to claim 1, wherein said least one portion of said representation of said object comprises a window having a substantially narrow width and a length substantially longer than said identifying feature in a corresponding direction in said representation of said object.

5. A method according to claim 1, wherein said at least one characteristic includes a total quantity of edges in said sampled portion of said representation at a predefined position in said representation, and wherein said step of locating said at least one feature includes the substep of:

comparing said total quantity of edges detected in said sampled portion with a predefined range of quantity values representative of a quantity of edges in said identifying feature.

6. A method according to claim 1, wherein said second representation consists of a sequence of n values, S_l to S_n, and said step of normalizing said at least one portion of said second representation of said at least one other characteristic into a range from a beginning value, B to an ending value E, includes the substeps of:

assigning a beginning value, B, to a first value, V_l of said sequence of values, assigning an ending value, E to a last value, V_n of said sequence of values, and assigning each intermediate value, V_i, in said sequence of values, a value according to the equation;

space="preserve" listing-type="equation">V.sub.i =B+((E-B)H((S.sub.i -S.sub.l)/(S.sub.n -S.sub.l))).

7. A method according to claim 6, wherein said comparing step includes the substeps of:

comparing each value, V_i, of said n values of said normalized sequence to a corresponding value of a predefined sequence of values representing said identifying feature;

rejecting said object if any of said values, V_i is greater than or less than a predefined tolerance of said corresponding value of said predefined sequence; and

not rejecting said object if all of said values, V_i are within said predefined tolerance of said corresponding value of said predefined sequence.

8. An apparatus for identifying an object having an identifying feature comprising:

means for creating a representation of said object;

means for locating at least one feature in said representation of said object having a corresponding characteristic similar to said at least one characteristic of said identifying feature and means for creating a second representation of at least one other characteristic of said located feature, said means for locating further including means for sampling at least one portion of said representation of said object at a predefined position in said representation, said means for sampling including means for sampling at predetermined locations across said representation of said object in a predefined direction, said locations being spaced apart by a distance corresponding to at least one characteristic of said identifying feature, and means for creating said second representation of said sampled portion of said representation, said second representation including a representation of at least one other characteristic of said located feature;

means for normalizing at least a portion of said second representation; and

means for comparing said normalized second representation of said at least one other characteristic of said at least one feature to a predefined representation of a characteristic of said identifying feature.

9. An apparatus according to claim 8, wherein said distance is less than a dimension of said identifying feature in said predefined direction.

10. An apparatus according to claim 8, wherein said distance is less than one half a dimension of said identifying feature in said predefined direction.

11. An apparatus according to claim 8, wherein said least one portion of said representation of said object comprises a window having a substantially narrow width and a length substantially longer than said identifying feature in a corresponding direction in said representation of said object.

12. An apparatus according to claim 8, wherein said at least one characteristic includes a total quantity of edges in said sampled portion of said representation at a predefined position in said representation, and wherein said means for locating said at least one feature includes:

means for comparing said total quantity of edges detected in said sampled portion with a predefined range of quantity values representative of a quantity of edges in said identifying feature.

13. A method for identifying an object having a barcode, said barcode having a first predefined characteristic and a second predefined characteristic, said method comprising the steps of:

1) creating a representation of said object;

2) locating at least one feature in said representation of said object having a characteristic similar to said first predefined characteristic of said barcode, the step of locating including the step of sampling at predefined locations across said representation of said object in a predefined direction, said locations being spaced apart by a distance corresponding to at least one characteristic of said bar code;

3) creating a representation of a second characteristic of corresponding to said second predefined characteristic;

4) creating a normalized representation of said second characteristic from said representation of said second predefined characteristic; and

5) comparing said normalized representation of said second characteristic to said second predefined characteristic of said barcode.

14. A method according to claim 13, wherein said first predefined characteristic includes the number of edges in a representation of said barcode along a predefined path, and said second predefined characteristic includes a list of locations of said edges along said predefined path.

15. A method according to claim 13, wherein said first predefined characteristic includes the number of edges of the same polarity in a representation of said barcode along a predefined path, and said second predefined characteristic includes a list of locations of said edges along said predefined path.