Research On Compressed Sensing Algorithms And Applications

Based on signal sparse features, compressed sensing use few measurement data with muchsmaller than the Nyquist sampling rate to realize signal reconstruction, and can greatly reducesignal processing cost. Since the theory was proposed, it draws widely attention in academics andindustry.this paper mainly research on compressed sensing algorithms and its applications.The first chapter summarizes the framework、significance of compressed sensing. The secondchapter mainly discusses the basic principle of compressed sensing、 compressed sensingalgorithms、 compressed sensing with noise and analyzes the uniqueness of sparse solution.The third chapter mainly introduces the greed algorithms in compressed sensing: orthogonalmatching pursuit algorithm, Stagewise Orthogonal Matching Pursuit, regularization orthogonalmatching pursuit algorithm and so on. The performance of these algorithms is analyzed.Based on the image stucture of salt and pepper noise, the fourth chapter applies compressedsensing to denoising, gives a denoising method. Compared to traditional denoising methods, theexperimental results show that the method’s efficiency.Based on the sparsity of natual images, the fifth chapter researches the image recovery from lossdata using compressed sensing. The experimental results show that under large data loss cases, theimage recovery algorithm still gives good reconstruction results.