Finite Ridgelet Transform And Its Applications In Image Processing

A number of facts have approved that wavelet mainly adapts to the isotropic singularity object.but to the anisotropic singularity,such as the verge of image and the wirelike character, wavelet is not the best tool. It is the reason that a array of proceeding means based on wavelet ,like image proceeding,cause the blur of image edges and details.But these are the most important information of a signal. Since ridgelet is proposed,the problem is solved well.In the paper,we study the application of finite ridgelet transform in image processing and propose a algorithm of image denoising and two watermarking algorithms.The first algorithm is obtained by substituting the dyadic discrete ridgelet for ortho- discrete ridgelet[16]. One of two watermarking methods is obtained as follows. first,given a natural image,dyadic wavelet is performed for it and a modual image is got ,then it is partitioned into small blocks ,and each block is performed by FRIT, the most significant coefficients are selected, and eventually the watermark is embedded.The second watermarking algorithm embeds the binary image watermarking in the selected coefficients according to the secret key basing on JND model in the ridgelet domain. A correlation function is used in the watermark detection process in two watermarking methods.Finally,these algorithms verified the validity through mimic experiments.Because finite ridgelet transform brings"wrap-around"effect, wiener filter is used for reducing its influence in the paper and a better result is gotten.But it is one of author's next work how more availably to reduce the"wrap-around"effect.