Application Of Lifting Wavelet Transform And Its Hardware Implementation

Lifting scheme is a new method for constructing wavelets and performing wavelet transform based on wavelet theory in recent years. It has many advantages for example simple structure, fast calculation, in-place operation. At the same time, it can realize integer to integer wavelet transform and can reconstruct original signal accurately. So it has been extensively used in the field of signal process. Lifting scheme based on wavelet transform is also used as the core transforming technology in JPEG2000, a newest image compression standard.Firstly, one-dimensional signal wavelet threshold denoising based on wavelet transform is studied . According to the shortcomings of soft-and-hard threshold function in wavelet based threshold denoising method, a novel threshold function with neighbor dependency is proposed. Meanwhile, a new method for determining denoising threshold value is also presented. Simulation results show that the new threshold function has preferable effects in signal denoising.Then the problem of image fusion is discussed .In order to fuse a high-resolution panchromatic image and a low-resolution image effectively , a novel image fusion algorithm using lifting wavelet transform and the IHS transform is proposed. The proposed merging algorithm is done by adding edge fusion factor in different region, and the region is divided based on evolutionary algorithm. Experimental results demonstrate that the proposed method can achieve better fusion performance in combining and preserving spectral-spatial information.In order to put lifting wavelet transform into real-time signal process, both the wavelet filters adopted in JPEG2000 and a new filter LS97 are introduced. A matlab simulation is performed to test the compatibility between CDF97 wavelet and LS97 wavelet, and the results show that they almost have the same performance. Then the architecture of Le Gall 53 wavelet and LS97 wavelet are designed. To improve the percentage of hardware utilizing, Le Gall 53 wavelet is combined with LS97 wavelet, and just one signal can control the switch of the two filters. Then the proposed architectures are described in Verilog HDL, simulation and implementation are performed on Xilinx's ISE6.2i to verify the feasibility of the design.