Multiwavelets are developed in recent years. They are a new addition to the body of wavelet theory. They preserve many properties of single wavelets while overcoming limitations of single wavelets. In practice

they perfectly combine smoothness

short support

orthogonality

symmetry.ect. The existing algorithms of multiwavelets denoising only provide theories of denoising but there is no concrete performing method. Hence

multiwavelets are combined with HeurShrink denoising method in this paper. A feasible denoising method

called MDH

is proposed and the denoising performances of D4 scalar wavelet are compared with that of MDH. The experimental results show that MDH is superior to D4 scalar wavelet in denoising via HeurShrink.