Image Denoising Method Based On Directional Information

The directional information play an important role in the field of image processing. So, it is key to detect and use this directional information in the domain of image denoising. In order to that, this thesis studied two aspects:The first aspect is contributed to generate a new directional filterbanks based on DFT modulated filterbanks and its implementation in image processing. It is easy to generate this new directional filterbanks by tensor product of analysis filterbanks of DFT modulated filterbanks, and easy to extend the number of directional subbands. In the field of image denoising, it can reserve the information in the low frequency, and have remarkable denoising results.The other aspect is devoted to the a new image denoising algorithm via doubly local Wiener filtering with windows based on Stationary wavelet transform (SWT) with the Dual tree complex wavelet transform (DTCWT). Doubly local Wiener filtering with the elliptic windows and the adaptive windows determined by the directional subbands is used in the field of image denoising. Experimental results and analysis are given to demonstrate the validity of the proposed algorithm.