CT Object Location via Orthogonal Projection Synthesis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current CT imaging technologies face challenges in accurately locating objects, particularly metal objects, during the image reconstruction process, which can lead to inaccuracies and artifacts in three-dimensional space determination.
Innovation Solution
A method involving the computation of a first projection image using a projection synthesis algorithm, followed by the selection of two orthogonal projection images from cone beam projections, and determining object locations in both two-dimensional and three-dimensional spaces using center of gravity calculations, effectively minimizing overlapping regions and reducing calculation errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional CT reconstruction algorithms are used to locate objects, then the imaging process can be completed, but the location accuracy of objects (especially metal objects) deteriorates due to artifacts and overlapping projections
Solution Approach 1:
The patent segments the 3D space into multiple 2D projection planes (top view, front view, side view) and processes object localization on each plane separately. By dividing the complex 3D localization problem into simpler 2D sub-problems, the method avoids artifacts and overlapping issues that plague conventional 3D reconstruction, thereby improving both measurement precision and reliability of object location.
2Measurement precision
If iterative reconstruction steps are used to improve location accuracy, then measurement precision improves, but the processing time and computational complexity increase
Solution Approach 1:
The patent performs preliminary actions by first identifying object centers of gravity in each 2D projection plane before attempting 3D localization. This preliminary 2D localization step provides accurate initial positions that eliminate the need for time-consuming iterative reconstruction, achieving high measurement precision while minimizing processing time.
3Measurement precision
If multiple projection images are used to reduce overlapping regions, then object location accuracy improves, but the device complexity and computational load increase
Solution Approach 1:
The patent transitions from 3D space to multiple 2D projection planes (dimensionality reduction) to simplify the localization problem. By projecting 3D objects onto 2D planes where overlapping is minimized, the method achieves high location accuracy while reducing computational complexity compared to direct 3D processing of multiple projections.
Data Source
AI summary
The present disclosure provides methods and devices for locating a plurality of interested objects in CT imaging. Location of the interested objects in the three-dimensional space can be determined by using three projection images that are substantially perpendicular to each other. The method can rapidly locate interested objects in a CT image without pre-reconstruction of the CT image even if there are a plurality of interested objects in the field of view. The algorithm does not involve interactive steps. The method is rapid and effective, and thus applicable to industrial applications.