Target Attribute Assisted Compressive Sensing for Image Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image reconstruction algorithms using compressive sensing face challenges with lower reconstruction speed, lower reconstruction precision, and excessive dependence on sparsity, particularly when the sparsity of a small target image signal is unknown, and fail to effectively utilize natural characteristics of image signals.
Innovation Solution
A method and system for image reconstruction using target attribute assisted compressive sensing, which involves initializing parameters, partitioning the image space into sub-blocks based on target attributes, updating the atom set, and performing least square estimation to efficiently select the dictionary subspace, thereby reducing the number of iterations and improving reconstruction quality without relying on sparsity or intra-value correlation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If subspace OMP algorithms are used for image reconstruction, then reconstruction precision is improved, but reconstruction speed deteriorates due to heavy computation burden
Solution Approach 1:
The patent segments the image space into multiple sub-blocks based on target attributes (size, position, shape). Each sub-block is processed independently with its own dictionary subspace, dividing the large-scale reconstruction problem into smaller, more manageable parts that reduce overall computation burden while maintaining reconstruction precision.
Solution Approach 2:
The patent applies local quality by using target attribute information to create different dictionary subspaces for different regions of the image. Each sub-block has a tailored dictionary subspace that matches the local characteristics, improving reconstruction precision in each region while avoiding the need to process the entire image with a uniform, computationally expensive approach.
2Productivity
If fixed subspace algorithms are used, then reconstruction speed is improved by reducing iterations, but reconstruction precision deteriorates due to excessive dependence on sparsity estimation
Solution Approach 1:
The patent performs preliminary action by using target attribute information (size, position, shape) to pre-construct dictionary subspaces before the main reconstruction process. This preliminary step based on prior knowledge of target characteristics enables faster convergence while maintaining high reconstruction precision, eliminating the need for extensive sparsity estimation iterations.
Solution Approach 2:
The patent introduces target attribute information as an intermediary that bridges the gap between speed and precision. This intermediary information guides the selection and construction of dictionary subspaces, enabling the algorithm to converge faster while maintaining high reconstruction accuracy without relying on trial-and-error sparsity estimation.
3Productivity
If dictionary subspace is selected based on intra-value correlation, then algorithm convergence is improved, but reconstruction precision deteriorates by ignoring natural characteristics of image signals
Solution Approach 1:
The patent changes the parameters used for dictionary subspace selection from generic intra-value correlation metrics to target attribute parameters (size, position, shape). This parameter change enables the algorithm to leverage natural characteristics of image signals containing targets, improving reconstruction precision while maintaining efficient convergence through the guidance of meaningful target-based parameters.
Data Source
AI summary
The present invention provides a method for image reconstruction using target attribute assisted compressive sensing, including an initializing step, a subspace partitioning step, an atom set updating step, a sparse coefficient updating step and an outputting step. The present invention further provides a system for image reconstruction using target attribute assisted compressive sensing. A technical scheme provided by the present invention will introduce auxiliary information capable of reflecting target features into subspace partitioning in a case of unknown sparseness of a small target image signal, thereby accurately and effectively selecting the most closely matching dictionary subspace, and realizing efficient and rapid reconstruction of the small target image signal.