Research On Wavelet De-noising Algorithm And The Design Of Its Soft IP Core

Wavelet theory, as being studied much more heatedly nowadays, has become a significant signal processing tool. De-nosing is one major branch of its application. Generally speaking, there are two wavelet de-nosing approaches, one's named'modulus maxima de-noising'which is proposed in the 1990s. However, the large computation amount of this algorithm limits its uses. Therefore, threshold de-noising is discussed more in the paper. Three keystones of threshold de-noising are: decompose level number estimate, shrinkage estimate and shrinking function select. Those three parts are discussed from several points of view in the paper and the simulations are given in the end.Compared to hardware, software can not be satisfactory when facing real time requirement. To solve this problem, the design of FPGA soft IP core of wavelet threshold de-noising is proposed thereafter. The thought and structure of soft IP core designing based on which the soft IP is built, is given.In the end, conclusion and expectation for future discussion are given.