2016 International Conference on Mechanical, Electronic and Information Technology Engineering (ICMITE 2016)

74 / 2015-12-19 16:16:32

Hybrid denoising algorithm based on wavelet decomposition and classification filtering

5685,8832,8833,8834,8835

Noise reduction is a widespread problem in image processing field. In this paper, a kind of image denoising method based on wavelet decomposition and classification filtering is proposed. First, use wavelet thresholding denoising method to remove the noise in the high frequency part of the image, then use classification filtering denoising algorithm to remove the noise in the low frequency approximation part part of the image, then use high frequency wavelet coefficients after threshold processing to reconstruction of high frequency part of the image, at last, combined with low frequency part and high frequency part image to get final noise reduction image. This algorithm not only can deal with gauss noise, but also can reduce salt and pepper noise, and gauss noise and salt and pepper mixed noise. This algorithm both can be used for grayscale image denoising and RGB colour image denoising. This algorithm structure is simple and easy to implement. Experiments show that this algorithm has good performance in denoising, noise reduction image is clarity.