Image denoising by preserving geometric components based on weighted bilateral filter and curvelet transform

Preservation of geometric components during image denoising using weighted bilateral filter and curvelet transforms is explored in this research. The proposed method emphases the texture and artifacts in an image while removing noise efficiently. Restoration of these details in an image not only improves the quality of image but also provides certain intelligence to the user for image understanding. Here, high frequency components are separated through weighted bilateral filter undergo curvelet transforms which leads to retaining of geometric features during the removal of noise components. Based on this, we propose a new method known as WBFCT and tested the performance in a simulated environment. Through a series of simulation of experiments we have compared the denoising performance of WBFCT with Standard Bilateral Filter (SBF), Robust Bilateral Filter (RBF), Weighted Bilateral Filter (WBF), LPG-PCA, KSVD, Curvelet only (Curvelet transform only without taking WBF), Wiener + Curvelet (Wiener filter in place of WBF), WBF + Wavelet (Wavelet transform in place of curvelet transform). Finally, the experimental outcomes divulged that present method has superior performance as compared to existing state-of-the-art methods pertaining to Gaussian noise.