An Improved Method for Colour Medical Image Compression Using Wrapping Based Hybrid Curvelet Transform

Abstract

Medical images have been acquired from patients in day to day life and are used for surgical and diagnostic purposes. They are stored as colourimages which normally consume high memory space. Hence, the compression of colour medical images is needed for storage and transmission. Due to the requirement of better accuracy in reconstruction and also loss of data is not tolerable, effective lossless compression algorithm is preferred. Colourmedical modalities are compressed using wrapping based curvelet transform hybrid by lifting, coded by Huffmanncoding and decompression using inverse transforms result in better Peak SNR and Compression value. This paper deals with the performance improvement such as improving the PSNR and CR in colourmedical image compression. Colourmedical modalities like Magnetic Resonance Image (MRI),Ultrasound, Computerised Tomography (CT)are compressed having varied image sizes and then result analysis is executed. It also ensures that other parameters like bits per pixel, correlation values, mean square error and difference in average are not getting affected and do not degrade the performance of compression/decompression. Most importantly it outperforms DCT, Haar, Contourlet and ripplet transforms with different coding.