Medical image de-noising schemes using wavelet threshold techniques with various noises

Nowadays the medical image processing has a vital role in diagnosis a human health in most of the health care institutions. Recently intensive research have been done in ultra sound imaging application to remove the blurry of noise image which is affected by random noise during the acquisition, analyzing and transmission. The blurry image produces inaccurate results that are difficult for doctors and biomedical engineers to extract fine physical exam report for patient. The discrete wavelet transform respectively uses the low pass filter and high pass filter to improve the affected image and getting lower and higher frequencies content image. The Discrete Wavelet Transform (DWT) has been applied for ultra sound image as affective tool in order to decompose the original image into details and approximation coefficients. It can be done by passing through it during the filters and reconstruct a sub-band details from the wavelet coefficients without changing the important features of the original image. This experimental work has been applied here to observe the medical image de-noising performance by using wavelet filter parameters such as PSNR & MSE from numerical results for the sake of efficient de-noising of noisy medical image. However, bayes threshold and Poisson noise techniques have been added to original image that produced maximum value of PSNR and minimum value MSE. The wavelet based de-noising algorithm has been investigated for medical image de-noising and best results of bayes technique and Hard & Soft threshold methods were achieved when different noises have been applied such as Poisson noises, Gaussian, Salt & Pepper and speckle. Meanwhile comparison evolution is being performed by taking individual noise values.