Research On The Single-pixel Imaging Methods Based On New Relationships Among The Coefficients And Asymmetric Measurement

Imaging technology?Carema?is one of the most important way to sense and record the information of the practical object with the visual sense organs.Since from the first practical silver analog camera produced by Daguerre in France in 1839 to the digtal camera born in 70s of the 20^th century,camera has been one of the indispensable devices in the daily life,research,military ffairs and security monitoring and control.To improve the imaging qulity under the dark environment and reduce the volume of the imaging system,the single-pixel imaging?SPI?teachnology is proposed by the researchers.SPI utilizes the measurement matrices designed by the users and the spatial light modulator?SLM?to generate a series of patterns,which measure the target object M times in sequence and obtains M meansurements with a single-pixel detector.The measurement matrices and M meansurements are adopted to reconstruct the target object image.In SPI,the light flux at the signle-pixel detector is higher and the detection effeciency is higher.This is suitable to the dark imaging environment.The volume is smaller and the cost of the system is lower.SPI can also combines the compressive sampling and the imaging technologies,which achieves compression and imaging simultaneously.This can reduce the waste of the sampling resource and size of transmitted data and achieves a broad application prospect.There are not only higher research significance,but also stronger potential application value in the research of single-pixel imaging.Superior imaging quality,less sampling resource,lower computational complexity and higher security are the main goals of single-pixel imaging.These also are significant problems when single-pixel imaging goes towards practical applications.To solve these significant problems,this thesis chooses the imaging quality as the pointcut,utilizes the prior knowledges,such as the characters of Gaussian white noise,relationships among the wavelet coefficients,sparsity of the target object images,and adopts the methods of analysis of signal-to-noise ratio in the measurements,comparsion between the computational complexity at the high and low resolutions,computation method of the significant wavelet coefficients,and the generation method of measurement matrices with the chaos system.The research on the noise suppression,lowering the reconstruction complexity,reducing the waste of sampling resource,and improving the security are implemented.We proposed some new ideas and methods,which also are proved by the simulation results.The main work and research contents are summarized as follows.1.An asymmetric compressive single-pixel imaging method is proposed.We defined the forward and backward noises in the sinle-pixel imaging system.With the signal-to-noise ratio and the characters of Gaussian white noise,we analyzed the methods to suppress the forward and backward noises and used the asymmetric measurement matrix to suppress the noises.The advantages of the proposed method are proved by the simulations.The best asymmetric ratio is also analyzed under different imaging environments.2.This thesis proposes an adaptive single-pixel imaging method based on continuous differencing operation and aggregated sampling methods.Based on the relationships among the coefficients of higher resolutions and lower resolutions,aggregated sampling is used to sample the significant coefficients directly,which reduces the reconstruction resolution and computational complexity.To reduce the waste of the information in measurement values and light flux,a general type of complementary measurement matrices generated with continuous differencing operation is proposed to replace the sparse 0/1 binary measurement matrices to reconstruct the significant coefficients.Based on the prior knowledges of the robustness of the coefficients of higher resolutions and lower resolutions against the noise,a ratio factor of significant coefficients is proposed to distribute the ratio of significant coefficients at different scales.This factor can recuce the human intervation in parameters setting and improve the rationality of parameters setting.The simulation reuslts under different imaging environments prove that the proposed method can reduce the reconstruction complexity effectively and the parameters setting is more reasonable.3.By introdcing the inter-scale and inter-scale sharing mechanisms and the fellow relationship among the wavelet coefficients,this thesis proposes an adaptive sensing method based on sharing mechanism and fellow relationship.The waste of sampling resource is reduced by localizing and sampling the significant coefficients using the sampled parent coefficients adequently.Simulation results show that the proposed method can achieve better imaging quality with lower sampling rate under different imaging environments.The sampling rate is lower for the same quality.4.This thesis proposes an adaptive single-pixel imaging method based on guided coefficients.The parent-children relationship and sharing mechanism are used adequently to localize the parent coefficients,which are not sampled yet and of which the children coefficients are significant.These parent coefficients are adopted as guided coefficients to guide the localization and sampling of significant coefficients.This can improve the localization accuracy.At last,the composite reconstruction method utilized the significant guided coefficients and their significant children coefficients combined with the ordinary significant coefficinets to improve the reconstruction quality.The simulation results prove that the proposed method improves the imaging quality and localization accuracy of significant coefficients with lower waste of sampling resource.5.This thesis proposes a secure simngle-pixel imaging method based on 2D Logistic map.A 2D Logistic map combined with its key and the asymmetric ratio is utilized to generate the asymmetric measurement matrices and encrypt the measurements.Compared with the exisitng secure SPI methods,the proposed method achieves the same security and the asymmetric measurement matrices supress the noise better with a better imaging quality.The key of the choas system combined with the asymmetric ratio in the asymmetric measurement matrices are transmitted in the secret channel instead of the measurement matrices.This strategy reduces the size of transmitted data.Simulation results show that the proposed method achieves advantages in imaging quality,security,size of transmitted data and computational complexity.