CT Scanner Edge Detection Using Sinogram Rebinning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing tomographic systems face challenges in accurately determining the physical extents and positioning of objects, especially irregularly shaped ones, within the scanning device, leading to confusion between external and internal measurements, and inefficiencies in data partitioning and image reconstruction.
Innovation Solution
A computed tomography (CT) system that continuously acquires raw data, rebins it into two-dimensional sinograms, determines the leading and trailing edges of objects, and reconstructs three-dimensional images using these sinograms, ensuring accurate edge detection and efficient data processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a light sensor array is used to identify leading and trailing edges of objects, then the determination speed is improved, but the accuracy deteriorates especially for irregularly shaped objects
Solution Approach 1:
The patent replaces the optical light sensor array with a computational approach using sinogram data processing. Instead of relying on physical light detection, the system uses mathematical transformations and algorithms to determine object edges from X-ray projection data, achieving both speed and accuracy.
Solution Approach 2:
The patent transforms the problem from direct spatial detection to parameter space analysis by converting object projection data into sinograms. By working with transformed parameters (sinogram values) rather than direct spatial coordinates, the system achieves more accurate edge determination while maintaining processing speed.
2Productivity
If objects are scanned continuously through the CT system, then productivity is improved, but measurement precision deteriorates due to repositioning confusion
Solution Approach 1:
The patent performs preliminary identification of objects and their extents before complete data acquisition. By detecting objects and determining their boundaries in advance using sinogram analysis, the system can properly partition data and maintain measurement precision throughout continuous scanning.
Solution Approach 2:
The system continuously monitors sinogram data to detect object presence and boundaries, providing feedback that triggers appropriate data acquisition and processing. This feedback mechanism ensures that data is captured and processed correctly for each object despite continuous movement through the scanner.
3Manufacturing precision
If data is partitioned into blocks for each object, then manufacturing precision is improved, but device complexity increases due to edge detection requirements
Solution Approach 1:
The patent uses a universal sinogram processing approach that handles multiple object detection and data partitioning tasks through a single unified algorithm. Instead of requiring separate complex systems for each function, the sinogram analysis serves multiple purposes: object identification, extent determination, and data partitioning simultaneously.
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 enables precise determination of object edges and boundaries, reduces processing time by reconstructing images only when objects are present, and improves the scanner's ability to keep pace with conveyor systems without frequent stops, while minimizing image detection errors.
Implementation Method 1
a detector configured to detect radiation emitted from the radiation source
Data Source
AI summary
A method for scanning a stream of objects includes continuously acquiring raw data of the stream of objects using an X-ray system, that includes a detector. The raw data of the stream of objects is rebinned into at least one two-dimensional sinogram. A leading edge and a trailing edge of a first object of the stream of objects is determined from the at least one two-dimensional sinogram and a three-dimensional image of the first object is reconstructed using the at least one two-dimensional sinogram.