A single-pixel terahertz wave imaging method and system

Abstract

A single-pixel imaging method and system based on spatial Fourier spectrum, using continuous pump light to irradiate silicon-based graphene to project a sinusoidal fringe mask, thereby obtaining a modulated terahertz image signal, after signal processing and solution The Fourier transform coefficients of the target image can be obtained by adjusting, and then only the inverse Fourier transform can be used to obtain the reconstruction of the target image. Furthermore, since the Fourier transform energy of natural scene images is concentrated in the low-frequency region, it is also possible to achieve under-sampling under the premise of ensuring image quality by discarding relevant information in the high-frequency region. Compared with the prior art, the present invention can greatly reduce the number of projection masks and shorten the sampling time under the condition of ensuring higher imaging quality with lower cost and system complexity.

Description

technical field [0001] The invention relates to the technical field of terahertz wave imaging, in particular to the imaging technology of single-pixel terahertz wave. Background technique [0002] Terahertz wave refers to (frequency 10 11 Hz-10 13 Hz or wavelength 30μm-3000μm) coherent electromagnetic radiation between the microwave band and the optical band. It has unique properties due to its special position in the transition from electronics to photonics in the electromagnetic spectrum. For example, the characteristic modes of many important biomolecules (such as proteins, DNA) and low-frequency vibrations of biological cells (such as the collective vibration and rotation of molecular skeletons, and the weak force between molecules) are in the range of the terahertz spectrum (spectral fingerprint properties). ). Based on terahertz spectral analysis, relevant information such as spatial conformation, reaction kinetics, hydration and biological functions of biomolecule...

AI Technical Summary

Problems solved by technology

[0006] 1. The sampling time is still long: due to the limitation of information requirements for compressed sensing algorithm reconstruction, there is a corresponding under-sampling sampling limit for target images with different complexity. If the sampling is lower than this limit, the image cannot be demodulated.

Generally speaking, the number of acquisitions needs to be greater than the image sparsity in order to reconstruct the target image well, and the image sparsity is related to the image complexity. The more complex the image, the more points need to be collected. In addition, the terahertz wave modulation depends on Modulator performance, and the current modulator performance is poor. In order to obtain higher image quality, the number of sampling points is required to be large, and it is difficult to directly reduce the number of sampling masks.

[0007] 2. High algorithm complexity: Due to the need to satisfy the constraint equidistant condition, the lower the correlation between the encoding matrix and the target image in the existing compressed sensing algorithm technology, the better. The complexity of the algorithm is also related to the projection mask. The limitation of undersampling makes the current algorithm complex. In addition, the algorithm complexity is exponentially related to the image resolution, which is not suitable for the reconstruction of high-resolution images.

[0008] 3. High system complexity: At present, the single-pixel terahertz wave imaging of the optical control modulator is the easiest solution to realize

However, this solution further increases the complexity and integration difficulty of the system, which is not conducive to the promotion of terahertz wave imaging

[0009] It can be seen that even if there are some methods in the existing technology that can reduce the imaging time or optimize the hardware system, they can only solve some technical problems, and there is still no complete set of methods suitable for the terahertz band that can realize Solutions with short sampling times, simple algorithms, and low system complexity and cost

Images