Method for inspecting defects in the shape of object

• US 6,895,108 B2
• Filed: 05/17/2002
• Issued: 05/17/2005
• Est. Priority Date: 11/05/1998
• Status: Expired due to Fees

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Kim, Myoung-jin
• Primary Examiner(s): AHMED, SAMIR ANWAR
• Application Number:
  US10/147,048
• Publication Number:
  US 20020191833A1
• Time in Patent Office: 1,096 Days
• Field of Search: 382/145, 382/147, 382/148, 382/149, 382/151, 348/87, 348/94, 348/95, 348/126, 348/129, 348/130, 356/390, 356/394, 356/237.4, 356/237.5, 438/7, 438/16, 250/559.39, 250/559.45, 250/559.46, 702/40, 702/82, 702/172
• US Class Current: 382/145
• CPC Class Codes: G01N 21/95607 using a comparative method

Abstract

A method of inspecting for defects in the shape of an object, using an inspected image obtained from an object to be inspected, comprising the steps of: setting a reference image of a reference model for determining the defect in shape of the object to be inspected; obtaining the imaged image from the object to be inspected; comparing the corresponding whole or local area of the reference image with the inspected image in terms of brightness; and determining as to whether or not the object is defective in shape. According to the present invention, the shape defect inspection does not rely upon the skilled person'"'"'s experience, can reduce time required for inspection, and can be made for the whole area of the inspection area.

Claims

1. A method of inspecting for defects in an object, comprising the steps of:

preparing a reference image representing at least a portion of a shape as a grayscale of unidirectionally varying unique values extending across a boundary of the shape within a range of tolerance accommodating occurrence of pixel noise;

obtaining an image from the object to be inspected;

making a comparison between an optical characteristic of corresponding areas of said reference image and the image obtained from the object; and

making a determination of whether said optical characteristic for said image obtained is within said range of tolerance of said corresponding area of said reference image.

2. The method of claim 1, with said comparison comprising a step of comparing brightness of a reference pixel of said image obtained with a corresponding pixel of said reference image.

3. The method of claim 1, further comprised of deriving from said reference image a minimum image and a maximum image that accommodate allowable errors in shape of the object by defining said range of tolerance.

4. The method of claim 3, further comprised of obtaining said reference image by:

obtaining an allowable brightness range of said reference pixel corresponding to an allowable position range of a border line of the object to be inspected, based on a function of variation in brightness of said reference pixel relative to the position of the border line; and

obtaining said minimum image and said maximum image respectively from a minimum value and a maximum value in the allowable range of brightness of said reference pixel.

5. The method according to claim 4, wherein the function considers an existence of a pixel noise.

6. The method according to claim 4, wherein an operation for the brightness comparison is represented in the following expression:

C(Q;

U,L)=π

[l(i,j)≦

q(i,j)≦

u(i,j)],

where;

l(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of the minimum image L,

q(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of the image Q, and

u(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of the maximum image U.

7. The method according to claim 1, further comprised of representing said image obtained and said reference image in grayscale.

8. The method of claim 1, further comprised of:

deriving from said reference image a minimum image and a maximum image for each of said corresponding areas of said reference image; and

establishing said range of tolerance in dependence upon said minimum image and said maximum image.

9. A method of inspecting for defects in an object, comprising the steps of:

preparing a reference image representing at least a portion of a shape as a grayscale of unidirectionally varying unique values extending across a salient feature of the shape within a range of tolerance accommodating occurrence of pixel noise;

establishing a minimum value and a maximum value of a characteristic common to each of a plurality of reference pixels forming said reference image;

obtaining an image from the object to be inspected;

making a comparison between a representative value of said characteristic for each of a plurality of inspection pixels forming said image obtained from the object, and said minimum value and said maximum value for a corresponding one of said reference pixels;

making a determination of whether the object to be inspected is within a range of manufacturing tolerance, in dependence upon said comparison.

10. The method of claim 9, further comprised of:

establishing said minimum value and said maximum value to gauge deviations between said salient feature and corresponding feature of the object;

using said minimum value and said maximum value to set said range of tolerance; and

rejecting the object whenever said determination establishes that a plurality of said representative values are outside said range of tolerance.

11. The method of claim 9, further comprised of:

setting a minimum value in correspondence with a minimum degree of brightness provided by said reference image; and

setting said maximum value in correspondence with a maximum degree of brightness provided by said reference image.

12. The method of claim 9, further comprised of:

deriving from said reference image a minimum image and a maximum image for each of said corresponding areas of said reference image; and

establishing said range of tolerance in dependence upon said minimum image and said maximum image.

13. The method of claim 9, further comprised of compensating for a presence of said pixel noise in said reference image by adjusting said minimum value and said maximum value.

14. The method of claim 9, further comprised of using brightness of said reference pixels and brightness of said inspection pixels as said optical characteristic.

15. The method of claim 9, further comprised of making said comparison in accordance with the following expression:

C(Q;

U,L)=π

[l(i,j)≦

q(i,J)≦

u(i,j)],

where;

(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of a minimum image L provided by said reference image,

q(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of the image Q provided by said inspection image, and

u(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of the maximum image U provided by said reference image.

16. The method of claim 10, further comprised of making said comparison in accordance with the following expression:

C(Q;

U,L)=π

[l(i,j)≦

q(i,j)≦

u(i,J)],

where;

l(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of a minimum image L provided by said reference image,

q(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of the image Q provided by said inspection image, and

u(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of the maximum image U provided by said reference image.

17. The method of claim 11, further comprised of making said comparison in accordance with the following expression:

C(Q;

U,L)=π

[l(i,j)≦

q(i,j)≦

u(i,J)],

where;

l(i,j) is said minimum degree of brightness obtained from a pixel positioned on the coordinates (i,j) of a minimum image L provided by said reference image,

q(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of the image Q provided by said inspection image, and

u(i,j) is said minimum degree of brightness obtained from a pixel positioned on the coordinates (i,j) of the maximum image U provided by said reference image.

18. A method of inspecting for defects in an object, comprising the steps of:

preparing a reference image representing at least a portion of a shape as a grayscale of values varying unidirectionally across a distinctive feature of the shape, between a minimum value and a maximum value of the grayscale;

establishing from said reference image, a range of tolerance accommodating occurrence of pixel noise between said minimum value and said maximum value;

obtaining an image from the object;

making a comparison in portions where the reference image and the inspection image mutually correspond, between a representation of said distinctive feature for each of a plurality of inspection pixels forming said image, and said range of tolerance; and

making a determination of whether the object meets manufacturing specifications, in dependence upon said comparison.

19. The method of claim 18, further comprised of

establishing said minimum value and said maximum value to gauge deviations between said salient features and corresponding features of the object;

using said minimum value and said maximum value to set said range of tolerance; and

rejecting the object whenever said determination establishes that a plurality of said representative values are outside said range of tolerance.

20. The method of claim 18, further comprised of making said comparison in accordance with the following expression:

C(Q;

U,L)=π

[l(i,j)≦

q(i,j)≦

u(i,j)],

where;

l(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of a minimum image L provided by said reference image,

q(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of the image Q provided by said inspection image, and

u(i,j) is a brightness value of a pixel positioned on the coordinates (i,j) of the maximum image U provided by said reference image.