Abstract: A threshold de-noising method based on dyadic wavelet transform is proposed here based on the wavelet shrinkage put forword by D.L. Donoho and I.M. Johnstone. To analyze the de-noising performance of this method, denoising experiments is conducted on the same image in different levels of Gaussian noise. The simulation results show that the threshold de-noising method based on dyadic wavelet transform not only suppresses the Gibbs phenomenon near the edge of the image effectively, but also it makes the peak signal to noise ratio of the de-noising images have a large extent improved at different noise levels, and have a very small range of fluctuation in different noise levels, which results indicate that de-noising performance of the threshold de-noising method based on dyadic wavelet transform has a strong stability.

Key words: image de-noising, dyadic wavelet transform, noise level, PSNR, performance analysis