Method of and system for classifying objects using histogram segment features of multi-energy computed tomography images
View Patent ↗A method of and a system for identifying objects using histogram segment features from multi-energy CT images are provided. The multi-energy CT images include a CT image, which approximates density measurements of scanned objects, and a Z image, which approximates effective atomic number measurements of scanned objects. The method comprises: computing a density histogram for each potential threat object; smoothing the density histogram using a low-pass filter; identifying peaks in the smoothed density histogram; assigning a segment to each peak; computing histogram segment features for each segment; classifying each potential threat object into a threat or a non-threat using computed features.
1. A method of classifying objects using volumetric multi-energy CT image data, wherein data from CT and Z (atomic number) images including data from at least one segmented potential threat object are provided, comprising:
A. Computing a density histogram for the potential threat object using the CT image data;
B. Dividing the density histogram of the potential threat object data into one or more segments;
C. Computing the histogram segment features for each segment using the CT and Z image data;
D. Classifying each potential threat object into a threat or a non-threat using the computed histogram segment features.
2. The method of claim 1 , wherein Step B includes applying a low-pass filter to the density histogram.
3. The method of claim 1 , wherein Step B includes identifying peaks in the density histogram.
4. The method of claim 1 , wherein Step B includes identifying a starting point of each segment in the density histogram.
5. The method of claim 1 , wherein Step B includes identifying an ending point of each segment in the density histogram.
6. The method of claim 3 , wherein Step B includes assigning a segment to each identified peak in the density histogram.
7. The method of claim 1 , wherein Step C includes computing the mean density for each segment.
8. The method of claim 1 , wherein Step C includes computing the peak density for each segment.
9. The method of claim 1 , wherein Step C includes computing the density standard deviation for each segment.
10. The method of claim 1 , wherein Step C includes computing the width for each segment.
11. The method of claim 1 , wherein Step C includes computing the skewness for each segment.
12. The method of claim 1 , wherein Step C includes computing the kurtosis for each segment.
13. The method of claim 1 , wherein Step C includes computing the mean atomic number for each segment.
14. The method of claim 1 , wherein Step C includes computing the Z standard deviation for each segment.
15. A system for classifying objects using data from volumetric multi-energy CT images, wherein data from CT image and Z (atomic number) images including at least one segmented potential threat object are provided, comprising:
A. A module arranged and configured so as to compute a density histogram for the potential threat object using CT image data;
B. A module arranged and configured so as to divide the density histogram of potential threat object into one or more segments;
C. A module arranged and configured so as to compute the histogram segment features for each segment using the CT and Z image data;
D. A module arranged and configured so as to classify the potential threat object into a threat or a non-threat using the computed histogram segment features.
16. The system of claim 15 , wherein Module B includes a unit arranged and configured so as to apply a low-pass filter to the density histogram.
17. The system of claim 15 , wherein Module B includes a unit arranged and configured so as to identify peaks in the density histogram.
18. The system of claim 15 , wherein Module B includes a unit arranged and configured so as to identify a starting point of each segment in the density histogram.
19. The system of claim 15 , wherein Module B includes a unit arranged and configured so as to identify an ending point of each segment in the density histogram.
20. The system of claim 15 , wherein Module B includes a unit arranged and configured so as to assign a segment to each identified peak in the density histogram.
21. The system of claim 15 , wherein Module C includes a unit arranged and configured so as to compute the mean density for each segment.
22. The system of claim 15 , wherein Module C includes a unit arranged and configured so as to compute the peak density for each segment.
23. The system of claim 15 , wherein Module C includes a unit arranged and configured so as to compute the density standard deviation for each segment.
24. The system of claim 15 , wherein Module C includes a unit arranged and configured so as to compute the width for each segment.
25. The system of claim 15 , wherein Module C includes a unit arranged and configured so as to compute the skewness for each segment.
26. The system of claim 15 , wherein Module C includes a unit arranged and configured so as to compute the kurtosis for each segment.
27. The system of claim 15 , wherein Module C includes a unit arranged and configured so as to compute the mean atomic number for each segment.
28. The system of claim 15 , wherein Module C includes a unit arranged and configured so as to compute the Z standard deviation for each segment.