The Research Of Terahertz Imaging Technology Based On Block Compressive Sensing Theory

The terahertz wave generally refers to the electromagnetic wave having a frequency in the range of 0.1 THz to 10 THz,which is located between the infrared frequency and microwave frequency.It has unique characteristics such as coherence,low energy,penetration,fingerprint characteristics,and transient characteristics.These characteristics makes it extremely promising in the fields of security,military,communications and other fields.However,the lack of devices related to terahertz range has greatly limited the development and promotion of terahertz imaging technology.As a new signal processing theory,compressive sensing theory can break the limitation of Nyquist's sampling law and accurately reconstruct the signal at a sampling rate of 25% or lower.Compared with traditional imaging technology,the imaging time can be greatly shortened.Applying the compressive sensing technology to terahertz imaging with using the single-pixel detector that is sensitive to terahertz range can effectively achieve fast imaging,the imaging method has a simple structure and low cost.However,as a new theory,compressive sensing still has its disadvantage in experimental imaging.Based on the previous research work,this thesis points out the shortcomings of the commonly used compressive sensing imaging method,and proposes a two-pixel terahertz wave compressive sensing imaging method.The practicality of the proposed method is verified by theory and experiment.Then,aiming at the low resolution of terahertz imaging,a wide-beam modulation imaging system based on block compressive sensing theory is proposed.On the one hand,the effectiveness of the method is verified theoretically and experimentally.On the other hand,the further application of block compressive sensing and the problems involved in this area are further explored.The main innovations of the thesis include:1.The two-pixel compressive sensing method which combined with transmission path and reflection path is used for imaging.Compared with the traditional single-pixel imaging method,the quality and anti-noise ability of the reconstructed image are improved.It provides a new idea for the application of compressive sensing imaging technology in the terahertz field.2.The optimal size of the measurement matrix in block compressive sensing imaging is determined by the light source energy distribution.The reconstruction of four different measurement matrices are studied by computer simulation.Then,the Bernoulli matrix is used instead of a Gaussian matrix to design a metal mask for optimal reconstruction.3.By adopting the method of wide-beam modulation sampling instead of single-pixel random sampling,the quality and stability of the reconstructed image by using block compressive sensing method are improved.It provides a reference for the further development of block compressive sensing.4.By using the image post-smoothing process instead of measurement matrix optimization and the method of block overlap,the boundary effect in the block compressive sensing imaging is processed and a preliminary exploration is carried out.It provides a new idea for the further improvement of the block compressive sensing theory.