Research On Quantization Method Based On Compressive Sensing

In the traditional process of data transmission, the sampling frequency mustconform to the Nyquist Law. In the transmission of digital image and video signal,meeting up with the theory leads to great cost in storage and transmission. In recentyears, a rising theory appears, which is compressive sensing. This new method offersnew ways in signal processing. It utilizes the property of the sparse signal to obtain thecomplete information by acquiring the small mount message of the linear projectiondirectly. Besides it can compress the information simultaneously when sampling byresolving the parse formulation to rebuild it. As a result, it makes sampling easier andprocessing of the information more efficient.Although compressed sensing can effectively reduce the size of the amount ofdata to quantify, but the observed data is still very large, quantification is stillcompression perceived an important step in data compression, need to conductin-depth research.Finally, it presents some achievements of quantization in the base of thecompressive sensing and their property. Second, uniform quantization is complied incompressive sensing. Some new methods of uniform quantization are raised base uponthe traditional methods. Their properties are tested by the simulation. Finally itpresents the segmenting uniform quantization which performs better in processing andbandwidth in the same PSNR over the tradition as its simulation results shows.