Granted Patent B2

US 7,352,892 · App. 10/392,948 · Granted Apr 1, 2008

Abstract

Claims (74)

1. A method for digitally reconstructing a surface of an object, comprising:

receiving image data including pixel values from one or more two-dimensional optical images of the object using at least one predetermined configuration of illuminators;

estimating surface gradient information for each pixel location according to a relationship between the at least one predetermined configuration of illuminators and the received image data, the surface gradient information representing approximate slopes at locations on the surface of the object corresponding to the pixel locations; and

determining surface height information defining a shape of the surface of the object in greater than two dimensions using said surface gradient information and noise information representing a noise distribution in the one or more optical images.

2. The method of claim 1 , wherein said determining comprises: utilizing a Bayesian process to calculate the surface height information.

3. The method of claim 1 , further comprising:

using the surface gradient information to reconstruct a 2.5-D image of the surface of the object.

4. The method of claim 1 , further comprising:

using the surface gradient information to reconstruct a three-dimensional image of the surface of the object.

5. A method for digitally reconstructing a surface of an object, comprising:

receiving image data including pixel values from one or more two-dimensional optical images of the object;

estimating surface gradient information for each pixel location from the image data, the surface gradient information representing approximate slopes at locations on the surface of the object corresponding to the pixel locations; and

determining surface height information defining a shape of the surface of the object in greater than two dimensions using said surface gradient information and noise information representing a noise distribution in the one or more optical images,

wherein said determining comprises: utilizing a Bayesian process to calculate the surface height information, and

wherein said utilizing comprises:

retrieving a pre-calculated gradient-to-height matrix; and

multiplying the surface gradient information with the gradient-to-height matrix to determine the surface height information.

6. The method of claim 5 , wherein said determining further comprises: determining the surface height information using a multi-resolution process.

7. The method of claim 6 , wherein said determining further comprises: partitioning the one or more images into original cells of at least 2×2 pixels;

determining first surface height information representing the relative heights within each of the original cells;

estimating additional surface gradient information among the original cells;

determining additional surface height information representing the relative heights between each of the original cells; and

combining the first and additional surface height information to produce the surface height information.

8. The method of claim 7 , further comprising:

combining the original cells into larger cells of at least 2×2 original cells; and

repeating said estimating additional surface gradient information and said determining additional surface height information for the larger cells.

9. A method for digitally reconstructing a surface of an object, comprising:

receiving image data including pixel values from one or more two-dimensional optical images of the object;

estimating surface gradient information for each pixel location from the image data, the surface gradient information representing approximate slopes at locations on the surface of the object corresponding to the pixel locations;

determining surface height information defining a shape of the surface of the object in greater than two dimensions using said surface gradient information and noise information representing a noise distribution in the one or more optical images; and

determining an error distribution for each pixel location based on the estimated surface gradient information.

10. The method of claim 9 , wherein said determining the error distribution comprises:

creating a confidence map based on the estimated surface gradient information, the confidence map containing a confidence value for each pixel location indicating the reliability of the surface gradient information at that pixel location; and

applying a prior distribution to each pixel location based on the respective confidence value, the prior distribution having a variance value varying inversely to the confidence value.

11. The method of claim 10 , wherein said creating the confidence map comprises:

using design specification data of the object to determine the confidence value at each pixel location.

12. A method for digitally reconstructing a surface of an object, comprising:

receiving image data including pixel values from one or more optical images of the object using at least one predetermined configuration of illuminators;

estimating surface gradient information for each pixel location according to a relationship between the at least one predetermined configuration of illuminators and the received image data, the surface gradient information representing approximate slopes at spatial locations on the surface of the object corresponding to the pixel locations; and

determining surface height information defining a shape of the surface of the object in greater than two dimensions using noise information representing a noise distribution in the one or more optical images and multi-resolution processing of the surface gradient information.

13. The method of claim 12 , wherein said determining comprises:

estimating first surface gradient information for each pixel location from the image data;

determining first surface height information representing the relative heights within each of a plurality of sections of the pixel locations using a portion of the first surface gradient information associated with the respective sections of the pixel locations;

estimating additional surface gradient information among the sections;

determining additional surface height information representing the relative heights among each of the sections using the additional surface gradient information; and

combining the first and additional surface height information to produce the surface height information.

14. The method of claim 13 , further comprising:

combining the sections into larger sections including two or more of the sections; and

repeating said estimating additional surface gradient information and said determining additional surface height information for the larger sections.

15. The method of claim 13 , wherein at least one of said determining the first surface height information and said determining the additional surface height information comprises:

utilizing a multi-resolution Bayesian process to determine the surface height information.

16. The method of claim 14 , wherein at least one of said determining the first surface height information and said determining the additional surface height information comprises:

retrieving a pre-calculated gradient-to-height matrix representing the noise distribution of the surface gradient information within the one or more images; and

multiplying the surface gradient information with the gradient-to-height matrix to determine the surface height information.

17. The method of claim 13 , wherein at least one of said determining the first surface height information and said determining the additional surface height information comprises:

utilizing a wavelet process to determine the surface height information.

18. The method of claim 17 , wherein at least one of said determining the first surface height information and said determining the additional surface height information comprises:

calculating wavelet coefficients for the surface height information using the surface gradient information.

19. The method of claim 18 , wherein at least one of said estimating the first surface gradient information and said estimating the additional surface gradient information comprises:

subsampling the surface gradient information.

20. The method of claim 12 , further comprising:

using the surface gradient information to reconstruct a 2.5-D image of the surface of the object.

21. The method of claim 12 , further comprising:

using the surface gradient information to reconstruct a three-dimensional image of the surface of the object.

22. An optical inspection system, comprising:

an illumination source configured to provide illumination in at least one predetermined configuration of illuminators;

a surface gradient processor connected to receive image data including pixel values from one or more optical images of an object using the illumination from the illumination source and configured to estimate surface gradient information for each pixel location according to a relationship between the at least one predetermined configuration of illuminators and the received image data, the surface gradient information representing approximate slopes at spatial locations on a surface of the object corresponding to the pixel locations; and

a reconstruction processor configured to determine surface height information defining a shape of the surface of the object in greater than two dimensions using said surface gradient information and noise information representing a noise distribution in the one or more optical images.

23. The optical inspection system of claim 22 , further comprising:

a memory for storing a pre-calculated gradient-to-height matrix, said reconstruction processor being configured to multiply the surface gradient information with the gradient-to-height matrix to determine the surface height information.

24. An optical inspection system, comprising:

an illumination source configured to provide illumination in at least one predetermined configuration of illuminators;

a surface gradient processor connected to receive image data including pixel values from one or more optical images of an object using the illumination from the illumination source and configured to estimate surface gradient information for each pixel location according to a relationship between the at least one predetermined configuration of illuminators and the received image data, the surface gradient information representing approximate slopes at spatial locations on a surface of the object corresponding to the pixel locations; and

a reconstruction processor configured to determine surface height information defining a shape of the surface of the object in greater than two dimensions using noise information representing a noise distribution in the one or more optical images and multi-resolution processing of said surface gradient information.

Assignments (9)