An improved algorithm of search for compressive sensing image recovery based on lp norm

Compressed sensing theory by developing a signal sparse features, under the condition of far less than the Nyquist sampling rate, the correct signal is acquired with random sampling the discrete samples, and then through the nonlinear reconstruction algorithm reconstruction signal of high probability. Compression sensing was applied to image processing have potential application value, and the reconstruction algorithm is a key technology of compression perception. In order to improve the existing compressed sensing image reconstruction algorithm based on 4p norm reconstruction precision and efficiency of algorithm, In view of the problem of Hesse matix is not positive definite matrix need much computing in Lagrange function Sequence Quadratic Programming (SQP) method. In this paper ,we propose an improved algorithm image recovery based on 4p norm compressive sensing by introduction of value function,revised Hesse matrix Sequence Quadratic Programming method and combining image block compressed sensing. Through under different sampling rate and reconstruction algorithm of image reconstruction effect is compared, the image reconstruction algorithm is verified by the experiments made on the reconstruction accuracy and the algorithm time balance. This algorithm in the image block compression on some piece of effect remains to be improved, but on the whole, improve the image reconstruction precision and computing time. Therefore, higher precIsion and faster image reconstruction algorithm for further study.