Study of Modified Non Subsampled Controurlet Transform

Enhancing Image Processing with Modified Non Subsampled Controurlet Transform

Abstract

A number of image processing tasks are efficiently carried out using the domain of invertible linear transformation. Various areas of application such as Image compressing, Image denoising, Image coding etc. are efficiently been carried out using different types transformation techniques. An effective transform capture the essence of a given signal with defined basis functions. The set of basic functions completely characterizes the transform and these sets can be redundant or non-redundant in nature. In the redundant approach, it is possible to improve the set of basis function so as the representation is more efficient in capturing signal behavior. In addition, redundant representations are generally more flexible and easier to design. In various applications such as denoising, image enhancement, contour detection etc. a redundant representation significantly outperforms the non-redundant approach.