The Application Of Wavelet Transform To Inverse Halftoning

Because of the features of Multi-Resolution Analyses (MRA) of the wavelet, inverse halftoning based on wavelet is well-suited. Wavelet transformation decomposition of the halftone image performs processing to preserve most of the details and remove the halftone noise.The choice of wavelet basis function and coefficient adjusting are two keys of inverse halftone. The characteristics about simultaneous, orthogonality, symmetry and short support that are offered by wavelet are very important to inverse halftoning. So the proposed method selects B-spline and GHM as wavelet basis function. The realization of inverse halftoning depends on coefficient adjusting. Based on difference about noise and details in various scales, the soft threshold and max-valued of gradient are proposed for adjusting coefficients. The continuous tone image was obtained by wavelet reconstruction. Inverse halftoning is achieved according to thinking in the proposed method and process of wavelet transformation.Lately the quality assessment and linear regression analysis are shown by experimental results.