Parametric Image Reconstruction for Limited-Angle X-Ray CT
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image reconstruction methods in baggage scanners face challenges in accurately detecting objects with complex shapes and materials, particularly in limited-angle x-ray computerized tomography, where convex hull approximation may not provide sufficient accuracy for concave shapes.
Innovation Solution
A parametric image reconstruction method using a system with multiple x-ray source and detector arrays to generate projection data, which determines tangential x-rays to form convex polygons for outer and inner boundary curves, and fits a parametric model with material-specific parameters to improve boundary curve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If convex hull approximation is used for object boundary reconstruction in limited-angle x-ray CT, then the reconstruction process is simplified, but the accuracy deteriorates for concave shapes
Solution Approach 1:
The patent segments the boundary curve reconstruction into two distinct parts: convex regions and concave regions. Convex regions are reconstructed using traditional convex hull methods, while concave regions are reconstructed using a parametric model that explicitly models the concave geometry. This segmentation allows each region to be handled with the most appropriate method, maintaining simplicity where applicable while achieving accuracy where needed.
Solution Approach 2:
The patent applies different reconstruction qualities to different parts of the boundary curve. For convex portions, a simpler reconstruction approach is used, while for concave portions, a more sophisticated parametric model with multiple parameters is employed. This local differentiation of quality ensures computational efficiency is maintained overall while achieving high accuracy specifically where the concave geometry requires it.
2Ease of operation
If traditional image reconstruction methods are used in baggage scanners, then the system is easier to operate, but the detection accuracy for complex shapes and materials deteriorates
Solution Approach 1:
The patent introduces a parametric model as an intermediary between the raw x-ray projection data and the final image reconstruction. This parametric model serves as a mediator that incorporates material-specific parameters and geometric constraints, allowing the reconstruction algorithm to work with a simplified representation that captures essential features while maintaining operational simplicity.
Solution Approach 2:
The patent utilizes parameter changes by representing the boundary curves and material properties through a set of parametric variables. By adjusting these parameters during the reconstruction process, the system can adapt to different object shapes and materials without requiring fundamental changes to the operational workflow, thus maintaining ease of operation while improving detection accuracy.
3Loss of time
If limited-angle x-ray scanning is used, then the scanning time is reduced, but the reconstruction accuracy for complex geometries deteriorates
Solution Approach 1:
The patent applies preliminary action by pre-defining the parametric model structure and material-specific parameters before the actual reconstruction process. This preliminary setup includes establishing the functional form of the boundary curves and selecting appropriate material parameters, which allows the limited-angle scanning data to be more effectively utilized during reconstruction, thereby improving accuracy without requiring additional scanning time.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the accuracy of image reconstruction by forming precise boundary curves and material identification, enabling effective detection of objects with complex shapes and materials, such as partially filled containers.
Implementation Method 1
A set of x-ray scanning arrays can contain one or more radiation sources that are configured to emit x-ray radiation towards an item under inspection, and an array of detectors on the opposite side of the item to detect the x-ray radiation that is not completely absorbed by the item
Implementation Method 2
detect the x-ray radiation that is not completely absorbed by the item
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Systems and methods for modeling are provided. The method can include acquiring scan data associated with a plurality of x-ray projections of an item. The method can further include determining at least one closed boundary curve associated with the item. For example, the method can determine a first maximum area based on the scan data and determine at least one edge of the first maximum area. The method can further generate a model of the item using the closed boundary curve and a first material specific parameter for a material within the closed boundary curve. The method can utilize the model to generate computed scan data, compare the computed scan data to the scan data, and determine a goodness of fit. The method can further adjust the model of the item by altering at least one of the closed boundary curve and the material specific parameter.