Tomography Image Generation Using Parallel Thread Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating tomography images using light are slow due to the sequential processing of depth scan operations, interference signal acquisition, and optimal pixel pattern determination, especially when multiple pixel patterns need to be adjusted, leading to slower process speeds.
Innovation Solution
A method and apparatus that parallel process depth scan operations, interference signal generation, and pixel pattern determination using multiple threads, employing a digital micromirror device (DMD) for spatial shift modulation and intensity comparison to quickly determine and adjust pixel patterns for generating high-resolution tomography images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If sequential processing is used for depth scan operations, interference signal acquisition, and pixel pattern determination, then processing accuracy is maintained, but processing speed deteriorates
Solution Approach 1:
The patent divides the processing workflow into three independent threads: depth scan operation thread, interference signal processing thread, and pixel pattern determination thread. Each thread handles specific tasks independently, allowing parallel execution to improve processing speed while maintaining modular complexity management.
Solution Approach 2:
The patent transitions from sequential one-dimensional processing to parallel multi-dimensional processing by implementing multiple concurrent threads. This dimensional change in processing architecture enables simultaneous execution of depth scanning, signal acquisition, and pattern determination without increasing overall system complexity.
2Measurement precision
If multiple pixel patterns are adjusted sequentially, then optimal pattern determination accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary parallel processing of interference signals and candidate tomography image generation while depth scanning is ongoing. This preliminary action allows the system to prepare multiple pixel pattern candidates simultaneously, reducing the time required for final optimal pattern determination without compromising accuracy.
Solution Approach 2:
The patent maintains continuous parallel processing across all stages: depth scan operations continue while interference signals are being processed, and candidate images are being generated. This continuous useful action eliminates idle time between processing stages, significantly reducing total processing time while maintaining determination accuracy through comprehensive pattern comparison.
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 significantly enhances the speed of generating tomography images by parallel processing, allowing for faster determination and application of optimal pixel patterns, thereby improving the efficiency of the tomography image generation process.
Implementation Method 1
employing a digital micromirror device (DMD) for spatial shift modulation
Implementation Method 2
generating a tomography image by using light... acquire an interference signal
Data Source
AI summary
Provided are a tomography image generating method and an apparatus for generating a tomography image. The method of generating a tomography image includes, in response to a depth scan operation performed on an object, generating a candidate tomography image by using an interference signal acquired by the performed depth scan operation, determining a pixel pattern by using the generated candidate tomography image; and when the depth scan operation performed on the object is completed, generating a final tomography image of the object by using a finally determined pixel pattern. The generating the candidate tomography image and the determining are parallel processed by at least one processor during the depth scan operation being repeatedly performed.