Comparison of Common Algorithms for Single-Pixel Imaging via Compressed Sensing

Wenjing Zhao,Lei Gao,Aiping Zhai,Dong Wang

DOI: https://doi.org/10.3390/s23104678

IF: 3.9

2023-05-11

Sensors

Abstract:Single-pixel imaging (SPI) uses a single-pixel detector instead of a detector array with a lot of pixels in traditional imaging techniques to realize two-dimensional or even multi-dimensional imaging. For SPI using compressed sensing, the target to be imaged is illuminated by a series of patterns with spatial resolution, and then the reflected or transmitted intensity is compressively sampled by the single-pixel detector to reconstruct the target image while breaking the limitation of the Nyquist sampling theorem. Recently, in the area of signal processing using compressed sensing, many measurement matrices as well as reconstruction algorithms have been proposed. It is necessary to explore the application of these methods in SPI. Therefore, this paper reviews the concept of compressive sensing SPI and summarizes the main measurement matrices and reconstruction algorithms in compressive sensing. Further, the performance of their applications in SPI through simulations and experiments is explored in detail, and then their advantages and disadvantages are summarized. Finally, the prospect of compressive sensing with SPI is discussed.

engineering, electrical & electronic,chemistry, analytical,instruments & instrumentation

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to select appropriate measurement matrices and reconstruction algorithms to optimize imaging performance when using Compressed Sensing (CS) techniques in Single - Pixel Imaging (SPI). Specifically, the paper focuses on:

1. **Concept Review and Summary**: Review the concept of Compressed Sensing SPI and summarize the main measurement matrices and reconstruction algorithms.
2. **Performance Exploration**: Explore in detail the application performance of these methods in SPI through simulation and experiments.
3. **Summary of Advantages and Disadvantages**: Summarize the advantages and disadvantages of mainstream measurement matrices and reconstruction algorithms in Compressed Sensing SPI.
4. **Prospect Discussion**: Discuss the future development prospects of Compressed Sensing SPI.

### Main Contributions of the Paper

- **Systematic Review**: Conduct a systematic review of the concept of Compressed Sensing SPI and summarize the main measurement matrices and reconstruction algorithms.
- **Performance Evaluation**: Evaluate in detail the performance of different measurement matrices and reconstruction algorithms through simulation and experiments.
- **Analysis of Advantages and Disadvantages**: Comprehensively analyze the advantages and disadvantages of various measurement matrices and reconstruction algorithms, providing guidance for practical applications.
- **Future Outlook**: Discuss the future research directions and development potential of Compressed Sensing SPI.

### Paper Structure

1. **Introduction**: Introduce the basic concepts of traditional imaging techniques and single - pixel imaging, as well as the application background of Compressed Sensing in SPI.
2. **Principle of Compressed Sensing SPI**: Describe in detail the working principle of Compressed Sensing SPI, including the importance of measurement matrices and reconstruction algorithms.
3. **Selection of Measurement Matrices**: Classify and introduce random measurement matrices, partially orthogonal measurement matrices, and semi - deterministic random measurement matrices, and analyze their respective advantages and disadvantages.
4. **Selection of Reconstruction Algorithms**: Classify and introduce convex optimization algorithms, greedy algorithms, non - convex optimization algorithms, Bregman distance minimization algorithms, and total variation minimization algorithms, and analyze their respective advantages and disadvantages.
5. **Performance Comparison**: Compare in detail the performance of different measurement matrices and reconstruction algorithms through simulation and experiments.
6. **Conclusion and Outlook**: Summarize the main work of the paper and discuss the future research directions of Compressed Sensing SPI.

### Key Formulas

- **Measurement Process**:
\[
y=\Phi x
\]
where \(\Phi\in\mathbb{R}^{M\times N}\) is the measurement matrix, \(x\in\mathbb{R}^{N\times1}\) is the one - dimensional representation of the target signal, and \(y\in\mathbb{R}^{M\times1}\) is the measurement result.

- **Sparse Representation**:
\[
y = \Phi x=\Phi\Psi s=A s
\]
where \(\Psi\in\mathbb{R}^{N\times N}\) is the sparse dictionary, \(A\in\mathbb{R}^{M\times N}\) is the joint matrix, and \(s\in\mathbb{R}^{N\times1}\) is the sparse vector.

- **Optimization Problem**:
\[
\min_{\hat{s}\in\mathbb{R}^n}\|\hat{s}\|_{l_0}\quad\text{subject to}\quad A\hat{s}=y
\]

- **RIP Condition**:
\[
(1 - \delta)\|s\|_2^2\leq\|A s\|_2^2\leq(1 + \delta)\|s\|_2^2
\]
where \(\delta\in(0, 1)\) is the Restricted Isometry Constant (RIC).

Through these contents, the paper aims to provide researchers and engineers with a comprehensive guide to help them select and apply appropriate Compressed Sensing techniques in single - pixel imaging.