Fast Single Pixel Imaging System Based On Difference

The traditional signal sampling based on the Shannon-Nyquist sampling law requires that the sampling rate must be greater than twice the highest rate of the signal to accurately reconstruct the signal.However,with the growing number of sensors around,the demand for signal sampling has doubled and the cost of signal storage,transmission,and processing is still too large to follow the Nyquist sampling law.In recent years,the compression sensing theory proposed by D.Donoho,E.Candes and T.Tao,a Chinese-American scientist,can break the traditional Shannon-Nyquist sampling law.The theory indicates that most signals in nature are sparseness or can be found under some sparse bases after sparsely transforming,and can use a large coefficient and position to express a complex signal,that is,can use low sampling rate to reconstruct high precision Raw signal.In image processing,a single pixel camera is a typical application of compression sensing,RICE University researchers successfully developed single-pixel imaging system based on this theory,which has a milestone-like significance.Compared with traditional cameras,single pixel camera has the characteristics of simple optical design,high compression ratio,wide application range and good encryption performance.However,the imaging quality of the existing single pixel camera system is still poor,and it is easy to be affected by the environmental noise.At the same time,the efficiency of the original image is very low,and the current single pixel system cannot replace the traditional camera.Both in the hardware structure and the restoration algorithm,there is a huge space for optimization.This paper introduces the compressed sensing theory,technical background and basic framework of compressed sensing theory in image processing applications,and the important step in the acquisition,processing and recovery in detail.At the same time introduces the classic single pixel imaging experiment of the University of RICE,summarizes the basic line of single pixel imaging experiments.Based on the experimental results of RICE,the advantages and disadvantages of classical single pixel imaging system are pointed out.Based on the deep understanding of compression perception theory and the analysis of classical single pixel imaging experiment of RICE University,the paper mainly completed the following two aspects of innovative work:(1)Aiming at the shortcomings of the existing single pixel imaging system,a fast single pixel imaging system based on differential is proposed.The system is optimized from the hardware structure and the software algorithm to the classical single pixel experiment system.That is: the hardware structure to increase the other with the same model to achieve a single pixel sensor differential measurement of the operation,reducing the disturbance caused by environmental noise and enhance the classical single pixel imaging system on environmental noise and ambient light changes in robustness;In the algorithm,a fast restoration algorithm based on vector decomposition of measurement value is proposed,which greatly shortens the time of original signal reconstruction.(2)The new hardware differential structure and the fast original signal restoration algorithm proposed in this paper are simulated by MATLAB environment.Compared with the existing hardware structure and restoration algorithm,the new system proposed in this paper can significantly enhance single pixel imaging The robustness of the system to ambient noise and the time required to recover the original signal can be significantly reduced.