Iterative Image Reconstruction via Dynamic Point Subset Refinement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image reconstruction methods, such as those used in radio interferometry and magnetic resonance imaging, face challenges in achieving high accuracy while maintaining reasonable computational requirements, particularly when dealing with large datasets and non-uniform image resolution.
Innovation Solution
The method involves iteratively modifying subsets of points based on detected signal features, progressively increasing the number of points at feature locations, and using approximate message passing algorithms over a factor graph to reconstruct images efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If iterative reconstruction techniques are used to improve image reconstruction accuracy, then manufacturing precision is improved, but use of energy increases due to higher computational requirements
Solution Approach 1:
The image reconstruction process is segmented into multiple iterations, where each iteration processes a refined subset of points. The point set is divided into active points (near detected features) and inactive points, with only active points being processed in detail. This segmentation allows the algorithm to achieve high reconstruction accuracy by concentrating computational effort on critical regions while reducing overall computational energy requirements through selective processing.
2Manufacturing precision
If the number of points in subsets is increased to improve image reconstruction accuracy, then manufacturing precision is improved, but device complexity increases
Solution Approach 1:
The subset of points used for reconstruction is made dynamic rather than static. In each iteration, the algorithm detects signal features in the current reconstruction and dynamically updates the point subset to include points near detected features while excluding distant points. This dynamic adaptation allows the algorithm to achieve high accuracy with varying point densities across iterations, reducing the need for uniformly high complexity throughout the entire reconstruction process.
3Use of energy by moving object
If traditional gridding and discrete Fourier transform methods are used to reduce computational requirements, then use of energy is reduced, but manufacturing precision deteriorates
Solution Approach 1:
The algorithm applies local quality by using different processing strategies for different regions of the image space. Near detected signal features, the algorithm uses dense point sampling and iterative refinement to achieve high local accuracy. In regions far from detected features, the algorithm uses sparser sampling and fewer iterations, reducing computational energy requirements. This localized approach to quality control allows the system to achieve high overall accuracy without uniformly high computational costs across the entire image.
Data Source
AI summary
The present invention is notably directed to computer-implemented methods and systems for recovering an image. Present methods comprise: accessing signal data representing signals; identifying subsets of points arranged so as to span a region of interest as current subsets of points; reconstructing an image based on current subsets of points, by combining signal data associated to the current subsets of points; detecting one or more signal features in a last image reconstructed; for each of the detected one or more signal features, modifying one or more subsets of the current subsets, so as to increase, for each of the modified one or more subsets, a relative number of points at a location of said each of the detected one or more signal features. The relative number of points of a given subset at a given location may be defined as the number of points of said given subset at the given location divided by the total number of points of said given subset, whereby new current subsets of points are obtained; and repeating the above steps of reconstructing, detecting and modifying, as necessary to obtain a reconstructed image that satisfies a given condition.