Image Enhancement Based on Anti-symmetrical Biorthogonal Wavelet Recontruction
CHI Jian-Nan,ZHANG Chuang,ZHANG Zhao-Hui,WANG Zhi-Liang
DOI: https://doi.org/10.3724/sp.j.1004.2010.00475
2010-01-01
Abstract:In this paper, image multi-scale edge detection based on anti-symmetrical biorthogonal wavelet reconstruction is given detailedly in theory, namely convolution operation property of the anti-symmetrical biorthogonal wavelet transform, is deduced and its differential operator function is analyzed. Then, an algorithm of wavelet decomposition in which multi-scale edge can be detected is put forward. Based on the above, an novel approach of image enhancement based on anti-symmetrical biorthogonal wavelet reconstruction is presented. Firstly, multi-scale decomposition of wavelet of image is done. Then, module value and phasic angle image are achieved through half-reconstruction, by which edge images of different scales are extracted. And pixels of half-reconstruction image corresponding to edge image are enhanced. Finally,reconstruction is done and image is enhanced. In the approach presented, image edges are enhanced in the process of reconstruction of wavelet tower data. This method gives a new idea for image denoising and enhancement. Experimental results show the validity of the approach.