Abstract:The imaging quality of Fourier single pixel imaging (FSI) is determined by the number of samples, therefore higher image quality requires more sampling times. Common low-frequency sampling methods compromise high-frequency signals to reduce the number of samples, but this will reduce the resolution of the reconstructed image. The recent adaptive sampling methods can reconstruct high-quality images with low sampling ratio, but the calculation of sampling trajectories is a time-consuming process. To improve imaging efficiency, we propose a fast sampling method based on intermediate frequency prediction for FSI. This method utilizes the continuity and directionality of important frequency distributions in the Fourier spectrum, according to which regions of important high-frequency components can be quickly predicted by firstly identifying important frequency components on the intermediate frequency ring. The simulation and experiment results demonstrated our method is 9 times faster than existing adaptive sampling methods in predicting sampling trajectories while maintaining similar image reconstruction quality for an image of 256 x 256 pixels under 5% sampling ratio. In addition, our method can efficiently sample important spectral components of images that has different frequency distributions, which proves that this method has the potential for wide applications.

Improving Imaging Quality of Real-time Fourier Single-pixel Imaging via Deep Learning

Alternative deep learning method for fast spatial-frequency shift imaging microscopy.

Optimizing the Quality of Fourier Single-Pixel Imaging Via Generative Adversarial Network

Deringing and Denoising in Extremely Under-Sampled Fourier Single Pixel Imaging

Compressed Sensing Fourier Single Pixel Imaging Algorithm Based on Joint Discrete Gradient and Non-Local Self-Similarity Priors

Fast High-Resolution Imaging Combining Deep Learning and Single-Pixel Imaging

Fast Fourier single-pixel imaging based on Sierra–Lite dithering algorithm

Adaptive Fourier single-pixel imaging sampling based on frequency coefficients prediction

Spatial Temporal Fourier Single-Pixel Imaging

Fast and High-Quality Single-Pixel Imaging.

Towards High-Resolution Undersampled Single-Pixel Imaging: a Neural Network Perspective

FSPI-R&D: Joint Reconstruction and Detection to Enhance the Object Detection Precision of Fourier Single-Pixel Imaging

On the use of deep learning for single-pixel imaging

High-resolution Iterative Reconstruction at Extremely Low Sampling Rate for Fourier Single-Pixel Imaging Via Diffusion Model.

S2O-FSPI: Fourier single pixel imaging via sampling strategy optimization

High Imaging Quality of Fourier Single Pixel Imaging Based on Generative Adversarial Networks at Low Sampling Rate

Deep learning based projector defocus compensation in single-pixel imaging.

High fidelity single-pixel imaging

Efficient Large-Scale Single-Pixel Imaging

Learning-based adaptive under-sampling for Fourier single-pixel imaging

Fourier Single-Pixel Imaging Based on Sampling Prediction from Intermediate Frequencies