Research On The Reconstruction Method Based On Compressive Sensing

With the rapid development of modern society,there will be a huge amount of data at all times.It is urgent for us to seek effective data processing techniques,methods and means.The superiority of the compressed sensing technology is fully applied to the data processing,which can not only solve the problem of large data processing,but also alleviate the pressure of the hardware facilities in the collection,storage and operation.In this paper,based on the relevant theory and literature,we improve the orthogonal matching pursuit algorithm of the compressed sensing theory and make a comparative analysis by MATLAB simulation.First,the theory of compressed sensing is tested by MATLAB,and the random Gauss measurement matrix is analyzed:the relationship between the number of measurements M and the accurate reconstruction probability,and the relationship between the num-ber of measurements M and Sparseness K.Second,matching pursuit(MP)algorithm and orthogonal matching pursuit(OMP)algorithm is proposed based on an improved OMP algorithm,OMP algorithm put a single atom into the selected set instead of mul-tiple atoms.Third,OMP algorithm and improved OMP algorithm respectively from two dimensions of one-dimensional sine wave superposition signal and two-dimensional im-age signal simulation experiments,using subjective and objective evaluation method to make comparison and analysis of experimental results.The experimental results show that compared with OMP algorithm,improved OMP algorithm has a small increase in performance,and the reconstruction effect is more satisfactory.