An Improved Descreening Algorithm for Color Halftone Image Based on Wavelet Decomposition

In this paper, a descreening algorithm based on wavelet is given. Wavelet decomposition is used to eliminate the halftone noise while preserving most of the original image contents. First, convert the color image to color space CMYK with the linear transformation. Then design the corresponding filter according to the direction of each channel after wavelet decomposition. Furthermore, optional nonlinear edge-preserving filter can be applied as a postprocessing to eliminate the sawtooth. Finally, three functions are used to evaluate the quality of image managed. Experiment results show that this method keep a higher definition while getting the better descreening result.