Methods for accelerating data acquisition in single-pixel imaging systems

This paper presents an overview of existing methods for accelerating data acquisition in single-pixel imaging systems, including single-pixel cameras and ghost imaging. Three key approaches to accelerating data acquisition are considered, namely: multiplexing, pattern optimization, and the use of feedback from the detector. The first approach, pattern multiplexing, allows parallel processing of data in different spectral, polarization, or temporal channels. The second approach, pattern optimization, aims to reduce the number of measurements without compromising image quality. The paper considers random, orthogonal (e.g., Hadamard and Fourier) and modified patterns to accelerate the data collection process. The third approach, the use of feedback from the detector, provides adaptive pattern correction, which increases the speed and accuracy of image reconstruction. This approach is particularly effective in combination with neural networks. The methods considered highlight the relevance of developing high-speed imaging systems that are used in remote sensing, medicine, and other fields. The combination of the approaches considered opens new opportunities for creating real-time imaging systems.