Rician noise education with svm and iterative bilateral filter in diffrent type of medical images

Abstract

Reduction of noise has always existed as a traditional conventional dilemma or dispute that occurs in digital image processing. The medical images (CT, X-RAY, MRI, PET scan and SPECT) consist of infinitesimal information regarding brain, heart, nerves and many more. Medical images may get depraved while transmission. When such images are depraved by different type of noise, it becomes impassable to recover a person from pernicious effects. Recent researches for de-noising applied the technique based on wavelet thresholding which proven to be promising because this technique is competent for suppressing the noise as well as preserving the every minutiae details of high frequency signal. Here, in this thesis a new technique based on Low Rank Matrix Decomposition (LRMD) using SVM has been prospected. The support vector machines have proved to achieve good generalization performance with no prior knowledge of the data. This technique will remove aliasing affect and noise from the noisy image. The aim of Low Rank Matrix approximation based upon image enhancement that it eliminates various types of noises from medical image. In digital image processing to denoise the medical image filters are mainly used to suppress either the high frequencies in the image that is smoothing the image or the lower frequencies that is enhancing or detecting edges in the image. The iterative bilateral filter is a non-linear filter and edge-preserving noise-reducing smoothing filter for medical images. The proposed approach expropriates the Additive Rician noise from the CT images of a patient and enhances the CT images quality. The proposed work comprises of three aspects namely; preprocessing, then training and concluded with testing. In the phase of preprocessing, input image afflicted by the Rician noise is transformed and in training phase, the obtained image is fed as input to the Low Rank Matrix Decomposition. Finally, testing phase, here the input image i.e. CT image is examined using SVM classifier for enhancing the quality of t