Image De-noising using 2-D Circular-Support Wavelet Transform
Images are often suffering from two main corruptions (unwanted modifications). These modifications in image accuracy are categorized as blur and noise. Noise appears during different image processing phases of acquisition, transmission, and retrieval. The purpose of any de-noising algorithm is to remove such noise while maintaining as much as possible image details. A 2-D circular-support wavelet transform (2-D CSWT) is anticipated in this paper to be utilized as an image de-noising algorithm. The realization of such de-noising algorithm is accomplished in the form of some competent mask filters. De-noising by thresholding processes can be applied on all 2-D high-pass coefficient channels with different thresholding levels. Lena noisy image with different levels of noise (Salt and Pepper, and Gaussian) has been used to assess the performance of such de-noising scheme. Test are done in terms of PSNR and correlation factor of the reconstructed image. A comparative study between the Conventional wavelet transform and the 2-D CSWT done in this paper.
- There are currently no refbacks.