Method and system for inspection of containers

1. A method for reconstructing and segmenting an image of the contents of a container, said method comprising:

(i) receiving actual scan data from a scan of the container;

(ii) reconstructing the actual scan data to obtain a reconstructed image;

(iii) segmenting the reconstructed image to obtain a segmented image, wherein said segmenting the reconstructed image comprises applying an algorithm including at least one of a seeded region-growing algorithm, an unseeded region-growing algorithm, a clustering algorithm, an edge detection algorithm, and a histogram-based algorithm to the reconstructed image;

(iv) deriving simulated scan data corresponding to the segmented image;

(v) calculating an error term based on a difference between the simulated scan data and the actual scan data;

(vi) determining whether a criterion is satisfied;

(vii) using the error term to produce a modified reconstructed image and repeating steps (iii) through (vii) with the modified reconstructed image substituted for the reconstructed image obtained in step (ii), if the criterion is not satisfied; and

(viii) outputting information from the segmented image if the criterion is satisfied.

2. A method in accordance with claim 1 , wherein said segmenting the reconstructed image to obtain a segmented image further comprises:

identifying a plurality of significant image elements in the reconstructed image;

grouping the plurality of significant image elements into a plurality of connected regions;

labeling each significant image element of the plurality of significant image elements that is grouped into at least a first connected region of the plurality of connected regions as a labeled image element of the first connected region;

calculating a representative value for the first connected region; and

creating the segmented image from the reconstructed image by replacing a value of each labeled image element of the first connected region with the representative value.

3. A method in accordance with claim 1 , wherein said segmenting the reconstructed image further comprises:

segmenting the reconstructed image to obtain at least one image segment;

calculating a representative image element value for the at least one image segment;

replacing an original value associated with each image element of a plurality of image elements within the at least one image segment with a value between the original value and the representative image element value; and

combining the plurality of image elements with a remainder of the reconstructed image to obtain the segmented image.

4. A method in accordance with claim 1 , wherein said reconstructing the actual scan data to obtain a reconstructed image comprises applying a filtered back-projection algorithm to the actual scan data.

5. A method in accordance with claim 1 , wherein said deriving simulated scan data corresponding to the segmented image comprises applying a forward-projection algorithm to the segmented image.

6. A method in accordance with claim 1 , wherein said using the error term to produce a modified reconstructed image comprises applying a filtered back-projection algorithm using the error term.

7. An inspection system for detecting objects within a container, said inspection system comprising a processor configured to:

(i) receive actual scan data from a scan of the container;

(ii) reconstruct the actual scan data to obtain a reconstructed image;

(iii) segment the reconstructed image to obtain a segmented image, wherein said processor is configured to segment the reconstructed image by applying an algorithm including at least one of a seeded region-growing algorithm, an unneeded region-growing algorithm, a clustering algorithm, an edge detection algorithm, and a histogram-based algorithm to the reconstructed image;

(iv) derive simulated scan data corresponding to the segmented image;

(v) calculate an error term based on a difference between the simulated scan data and the actual scan data;

(vi) determine whether a criterion is satisfied;

(vii) use the error term to produce a modified reconstructed image and repeat steps (iii) through (vii) with the modified reconstructed image substituted for the reconstructed image obtained in step (ii), if the criterion is not satisfied; and

(viii) output information from the segmented image if the criterion is satisfied.

8. An inspection system in accordance with claim 7 , wherein said processor is further configured to:

identify a plurality of significant image elements in the reconstructed image;

group the plurality of significant image elements into a plurality of connected regions;

label each significant image element of the plurality of significant image elements that is grouped into at least a first connected region of the plurality of connected regions as a labeled image element of the first connected region;

calculate a representative value for the first connected region; and

create the segmented image from the reconstructed image by replacing a value of each labeled image element of the first connected region with the representative value.

9. An inspection system in accordance with claim 7 , wherein said processor is further configured to:

segment the reconstructed image to obtain at least one image segment;

calculate a representative image element value for the at least one image segment;

replace an original value associated with each image element of a plurality of image elements within the at least one image segment with a value between the original value and the representative image element value; and

combine the plurality of image elements with a remainder of the reconstructed image to obtain the segmented image.

10. An inspection system in accordance with claim 7 , wherein said processor is further configured to apply a filtered back-projection algorithm to the actual scan data to obtain the reconstructed image.

11. An inspection system in accordance with claim 7 , wherein said processor is further configured to apply a forward-projection algorithm to the segmented image to derive simulated scan data corresponding to the segmented image.

12. An inspection system in accordance with claim 7 , wherein said processor is further configured to apply a filtered back-projection algorithm using the error term to produce the modified reconstructed image.

13. A computer-readable storage medium, said computer-readable storage medium being encoded with a computer program comprising a code segment that configures a processor to:

(i) receive actual scan data from a scan of a container by an inspection system;

(ii) reconstruct the actual scan data to obtain a reconstructed image;

(iii) segment the reconstructed image to obtain a segmented image, the code segment configuring the processor to segment the reconstructed image by applying an algorithm including at least one of a seeded region-growing algorithm, an unseeded region-growing algorithm, a clustering algorithm, an edge detection algorithm, and a histogram-based algorithm to the reconstructed image;

(iv) derive simulated scan data corresponding to the segmented image;

(v) calculate an error term based on a difference between the simulated scan data and the actual scan data;

(vi) determine whether a criterion is satisfied;

(vii) use the error term to produce a modified reconstructed image and repeat steps (iii) through (vii) with the modified reconstructed image substituted for the reconstructed image obtained in step (ii), if the criterion is not satisfied; and

(viii) output information from the segmented image if the criterion is satisfied.

14. A computer-readable storage medium in accordance with claim 13 , wherein the code segment further configures the processor to:

identify a plurality of significant image elements in the reconstructed image;

group the plurality of significant image elements into a plurality of connected regions;

label each significant image element of the plurality of significant image elements that is grouped into at least a first connected region of the plurality of connected regions as a labeled image element of the first connected region;

calculate a representative value for the first connected region; and

create the segmented image from the reconstructed image by replacing a value of each labeled image element of the first connected region with the representative value.

15. A computer readable storage medium in accordance with claim 13 , wherein the code segment further configures the processor to:

segment the reconstructed image to obtain at least one image segment;

calculate a representative image element value for the at least one image segment;

replace an original value associated with each image element of a plurality of image elements within the at least one image segment with a value between the original value and the representative image element value; and

combine the plurality of image elements with a remainder of the reconstructed image to obtain the segmented image.

16. A computer-readable storage medium in accordance with claim 13 , wherein the code segment further configures the processor to apply a forward-projection algorithm to the segmented image to derive simulated scan data corresponding to the segmented image.

17. A computer-readable storage medium in accordance with claim 13 , wherein the code segment further configures the processor to apply a filtered back-projection algorithm using the error term to produce the modified reconstructed image.