Enhanced design of embedded codec For medical imaging system

Abstract

Most of the software for specific tasks and applications are newlinemoving from Desktop computers to embedded systems The basic need newlinefor any embedded application is minimal computational complexity and newlineminimum storage requirement This leads to a large number of newlinecompression algorithms in the multimedia and related domains This newlinehas led to the need for more componentized software and the market is newlinelooking for off the shelf components The need for a higher newlinecompression ratio and reduced blocking artifacts the standards are newlineswitching from Discrete Cosine Transform to Discrete Wavelet newlineTransform Medical imaging systems require a loss less compression newlineModel In any system the probability of noise polluting the image is newlinehigh even in an embedded environment In this thesis an optimum newlinedouble filtering method is proposed for enhancement of medical newlineimages in the wavelet domain The experimentations took into newlineconsideration the embedded limitations and determined a static newlineoptimum window so as to achieve a high Peak Signal newlineto Noise Ratio for the reconstructed image for three different types of newlinenoise The thesis compares the double filtering method of enhancement newlinewith adaptive filtering for Gaussian noise newline newline