Electronic inspection system and methods of inspection

• US 4,817,184 A
• Filed: 04/14/1986
• Issued: 03/28/1989
• Est. Priority Date: 04/14/1986
• Status: Expired due to Term

• Current Assignee: Vartec Corp.
• Original Assignee: Vartec Corp.
• Inventors: Sklansky, Jack, Thomason, Robert L.
• Primary Examiner(s): Boudreau, Leo H.
• Application Number:
  US06/851,166
• Time in Patent Office: 1,079 Days
• Field of Search: 382/21, 382/22, 382/27, 382/8, 358/105
• US Class Current: 382/141
• CPC Class Codes: C12N 15/1003 Extracting or separating nucleic acids from biological samples, e.g. pure separation or isolation methods; Conditions, buffers or apparatuses therefor
• CPC Class Codes: C12N 15/117 Nucleic acids having immunomodulatory properties, e.g. containing CpG-motifs
• CPC Class Codes: C12N 2310/17 Immunomodulatory nucleic acids
• CPC Class Codes: G06T 2207/30141 Printed circuit board [PCB]
• CPC Class Codes: G06T 7/0004 Industrial image inspection

Abstract

An inspection system and method suitable for high speed inspection of articles having bounded surfaces are disclosed. The apparatus processes the image of the item being inspected to provide a first processed image comprising the modulus of the light intensity from the article being inspected, and a second modified image comprising, for each point, the angle of the highest light intensity gradient at that point in the image. The variability of the angle throughout the image is then determined, and the modulus and variability combined such as, by way of example, by point by point multiplication, to provide image data comprising a quantitative evaluation of the image. This data in turn may be combined into a single quantitative measure indicating the desired characteristics of the article being inspected for acceptance and rejection, or otherwise used to determine surface characteristics, locate known or unknown surface anomalies, and/or determine the size, shape, orientation, etc. of an outline or a portion thereof, as desired. The determination and use of the variability of the angle, either alone or in conjunction with the modulus highlights the boundaries of the object for emphasis or de-emphasis, as desired, as angle variability in the region of a boundary is normally very low. Also, other uses of the variability of the angle or other angle dependent information alone or with the modulus avoids the use of matrix rotations for nonoriented objects, again grossly speeding data reduction. Alternative embodiments and methods of data analysis, smoothing, clipping and shaping are disclosed.

Claims

1. A method of inspection of an article comprising the steps of

(a) digitizing the image of an article to be inspected to provide a first image array of first data values, each first data value being a digital representation of the intensity of the respective pixel of the image

(b) transforming the first image array to a second image array of second data values, each second data value indicating the direction of the intensity gradient for each respective first data value

(c) for each second data value of the second image array, determining a single third data value, each third data value being related to a respective second data value and responsive to the variability of the second data values within the neighborhood of that respective second data value, each neighborhood of each second data value being no more than a 3×

3 neighborhood and being chosen so that each second data value has the same relative neighborhood location as each other respective second data value has to its neighborhood location, whereby the second image array is transformed into a third image array, and

(d) characterizing the article being inspected utilizing, at least in part, the third data values.

2. The method of claim 1 further comprising the step of calculating, for each second data value based on its respective neighborhood of second data values, a mean second data value, and wherein, in step C, each third data value is responsive to the variability of the mean second data values within each corresponding neighborhood of mean second data values.

3. The method of claim 2 wherein each mean second data value is calculated for a corresponding neighborhood of second data values comprising an N by N array of second data values.

4. The method of claim 3 wherein N is two or more.

5. The method of claim 1 wherein in step C, each third data value is responsive to the number of different values of the second data values within the corresponding neighborhood of second data values.

6. The method of claim 1 further comprising the step of calculating, for each corresponding neighborhood of second data values, a mean second data value, and wherein, in step C, each third data value is responsive to the number of different values of the mean second data values within the corresponding neighborhood of mean second data values.

7. The method of claim 6 wherein each mean second data value is calculated for a corresponding neighborhood of second data values comprising an N by N array of second data values.

8. The method of claim 7 wherein N is two or more.

9. The method of claim 1 further comprised of the steps of

(e) calculating an array of fourth data values, each fourth data value being responsive to the magnitude of the intensity gradient at the respective image portion as indicated by the array of first data values

(f) combining each fourth data value with a respective third data value prior to step (d) of claim 1, whereby the article being inspected is characterized by the combination of the third and fourth data values.

10. The method of claim 9 wherein said array of fourth data values are smoothed prior to step (f).

11. The method of claim 9 wherein said fourth data values are transformed prior to step (f).

12. The method of claim 1 wherein each second data value is calculated from a corresponding N by N two dimensional array of first data values.

13. The method of claim 12 wherein N is at least 2.

14. The method of claim 12 wherein N is 3.

15. The method of claim 1 wherein said array of first data values are smoothed prior to step (b).

16. The method of claim 1 wherein said second data values are transformed prior to step (c).

17. The method of claim 1 wherein said second data values are transformed prior to step (c).

18. The method of claim 1 wherein said array of third data values are smoothed prior to step (d).

19. The method of claim 1 wherein said third data values are transformed prior to step (d).

20. The method of claim 1 wherein step C, each third data value is responsive to the number of distinct second data value pairs within the corresponding neighborhood of second data values which represent predetermined substantially opposing directions of intensity gradients.

21. A method of inspection of an article by analyzing the image thereof comprising the steps of

(a) digitizing on a line by line basis at least part of the image of an article to be inspected to provide a first image array of first data values, each first data value being a digital representation of the intensity of the respective element of the image

(b) transforming the first image array to a second image array of second data values, each second data value indicating the direction of the intensity gradient for each respective first data value

(c) for each second data value of the second image array, determining a single third data value, each third data value being related in relative array position to a respective second data value and responsive to the variability of the second data values within the neighborhood of that respective second data value, each neighborhood of each second data value being no larger than a 3×

3 neighborhood and chosen so that each second data value has the same relative neighborhood location as each respective second data value has its neighborhood location, whereby the second image array is transformed into a third image array, and

(d) characterizing the article being inspected utilizing, at least in part, the third data values.

22. The method of claim 21 wherein, in step (c), each third data value is responsive to the variability of the second data values within a 3×

3 neighborhood of that respective second data value.

23. The method of claim 22 wherein the 3×

3 neighborhood is a 3×

3 pixel neighborhood.

24. The method of claim 22 wherein the 3×

3 neighborhood is a 3×

3 neighborhood of cells, each cell being an N×

N array of pixels having a representative second data value for each cell.

25. Apparatus for the inspection of an article comprising camera means for viewing at least a part of an article to be inspected

digitizing means coupled to said camera means for digitizing on a line by line basis at least part of the image of an article to be inspected to provide an array of first data values, each first data value being a digital representation of the intensity of the respective pixel of the image

first calculating means coupled to said digitizing means for calculating an array of second data values, each second data value indicating the direction of the intensity gradient for each respective first data value

second calculating means coupled to said first calculating means for calculating an array of single third data values, each third data value being related in relative array position to a respective second data value and responsive to the variability of the second data values within he neighborhood of that respective second data value, each neighborhood of each second data value being no larger than 3×

3 neighborhood and chosen so that each second data value has the same relative neighborhood location as each other respective second data value has to its neighborhood location, whereby the second image array is transformed into a single third image array, and

characterizing means coupled to said second calculating means for characterizing the article being inspected utilizing, at least in part, the third data values.

26. The apparatus of claim 25 wherein said second calculating means is a means for calculating each third data value responsive to the variability of the second data values within a 3×

3 neighborhood of that respective second data value.

27. The apparatus of claim 26 wherein the 3×

3 neighborhood is a 3×

3 pixel neighborhood.

28. The apparatus of claim 26 wherein the 3×

3 neighborhood is a 3×

3 neighborhood of cells, each cell being an N×

N array of pixels having a representative second data value for each cell.

37. The apparatus of claim 28 wherein each representative second data value is a mean data value.

29. In a method of inspection wherein picture elements of an image of an article being inspected are analyzed to characterize the article, the improvement comprising:

(a) determining an array of data values, each data value representing the direction of the local intensity gradient for each picture element;

(b) determining from the direction of the local intensity gradient and as a single data value, the variability of the direction of the local intensity gradient within the neighborhood of each respective picture element, each neighborhood of each picture element being no more than a 3×

3 neighborhood and chosen so that each picture element has the same relative neighborhood location as each other picture element has to its neighborhood location; and

(c) characterizing the article being inspected utilizing, at least in part, the variability of the direction of the local intensity gradient within the neighborhood of a plurality of picture elements.

30. A method of inspection of an article comprising steps of

(a) digitizing the image of an article to be inspected to provide an array of first data values, each first data value being a digital representation of the intensity of the respective pixel of the image

(b) calculating an array of second data values, each second data value indicating the direction of the intensity gradient for each respective first data value

(c) calculating an array of third data values, each third data value being related to a respective second data value and responsive to the variability of the second data values within a local group of second data values forming the neighborhood of that respective second data value, each neighborhood of each second data value being chosen so that each second data value has the same relative neighborhood location as each other respective second data value has to its neighborhood location

(d) calculating an array of fourth data values, each fourth data value being responsive to the magnitude of the intensity gradient at the respective image portion as indicated by the array of first data values

(e) multiplying each fourth data value with a respective third data value, and

(f) characterizing the article being inspected by the product of the third and fourth data values.

31. The method of claim 30 wherein the article being inspected is characterized by counting the number of products of the third and fourth data values which are within a predetermined value range.

32. The method of claim 30 wherein at least one of said third and fourth data values are each transformed in a predetermined manner prior to combining respective ones of said third and fourth data values.

33. The method of claim 32 wherein the fourth data values are smoothed prior to combining with the third data values.

34. A method of inspection of an article by analyzing the image thereof comprising steps of

(a) digitizing on a line by line basis at least part of the image of an article to be inspected to provide an array of first data values, each first data value being a digital representative of the intensity of the respective element of the image

(b) calculating an array of second data values, each second data value indicating the direction of the intensity gradient for each respective first data value

(c) calculating an array of third data values, each third data value being related in relative array position to a respective second data value and responsive to the variability of the second data values within the 3×

3 neighborhood of that respective second data value, each neighborhood of each second data value being chosen so that each second data value has the same relative neighborhood location as each respective second data value has its neighborhood location, and

(d) characterizing the article being inspected utilizing, at least in part, the third data values,

wherein the 3×

3 neighborhood is a 3×

3 neighborhood of cells, each cell being an N×

N array of pixels having a representative second data value for each cell, and further comprising the step of calculating for each N×

N array of second data values a representative second data value, and wherein, in step C, each third data value is responsive to the variability of the representative second data values within each corresponding 3×

3 neighborhood of representative second data values.

35. The method of claim 34 wherein each representative second data value is a mean data value.

36. Apparatus for the inspection of an article comprising camera means for viewing at least a part of an article to be inspected

digitizing means coupled to said camera means for digitizing on a line by line basis at least in part of the image of an article to be inspected to provide an array of first data values, each first data value being a digital representation of the intensity of the respective pixel of the image

first calculating means coupled to said digitizing means for calculating an array of second data values, each second data value indicating the direction of the intensity gradient for each respective first data value

second calculating means coupled to said first calculating means for calculating an array of third data values, each third data value being related in relative array position to a respective second data value and responsive to the variability of the second data values within a 3×

3 neighborhood of that respective second data value, each neighborhood of each second data value being chosen so that each second data value has the same relative neighborhood location as each other respective second data value has to its neighborhood location,

characterizing means coupled to said second calculating means for characterizing the article being inspected utilizing, at least in part, the third data values,

wherein the 3×

3 neighborhood is a 3×

3 neighborhood of cells, each cell being an N×

N array of pixels having a representative second data value for each cell, and

further comprising third calculating means coupled to said second calculating means for calculating for each N×

N array of second data values a representative second data value, and wherein said second calculating means is coupled to said first calculating means through said third calculating means, whereby each third data value is responsive to the variability of the representative second data values within each corresponding 3×

3 neighborhood of representative second data values.