Integrating Local Pixel Grouping with Dual Tree Complex Wavelet Packets for De-Noising of Medical Images

摘要

Denoising of images is an important and rather challenging task, due to the peculiarities of the noise acquired by imaging sensors in ultrasounds (US), computer tomography (CT), or, of course, magnetic resonance images (MRI) and natural images. A large number of previous methods try to solve the denoising problem from a certain perspective, e.g., the wavelet and other techniques. In this paper, we propose a three-stage image denoising method by applying Dual Tree Complex Wavelet Packets (DTCWP) and LPG-PCA method. DTCWP and histon calculation are used as a method to identify the noisy pixel information and remove small amount of noise in first stage. The second stage yields an initial estimation of the image by removing most of the noise and the third stage is further refined from the output of the second stage. The three stages have the same process except for the parameter of noise level. As the noise is significantly brought down in the first and second stages, the procedure of morphological and region props matlab function tends to get further improved in the third stage leading to the astounding truth that the ultimate de-noising result is visually far superior. The suggested technique has been compared with our earlier de-noising method with Gaussian and salt-pepper noise. Simulation results say that in comparison with the other existing methods, the proposed approach has a better performance in de-noising and preserving image edges.