The Compressive Sensing Based on Biorthogonal Wavelet Basis

Compressive Sensing is a new theory to simultaneous sensing and compression. It enables a potentially large reduction in the sampling and computation costs for signals based on them having sparse representation in some basis. Many wavelet transformations were used in CS, such as Haar, db4, db6 and db8 wavelets. We test several wavelet basis from Daubechies and Biorthogonal family using Bayesian wavelet-tree structured CS. The Error Ratio between the original coefficient and the reconstructed coefficient, the PSNR of the original image and reconstructed image, and the Elapsed Time were used as the measurement indexes. Two images, Indor3 and Lena are used in experimental. The results indicate that the Biorthogonal wavelet family, from which bior2.8 and bior3.5 are used in this paper, can get the better results than other wavelet basis.