Performance Evaluation for Medical Images Using Partitioned Iterated Function Systems

Abstract

The compression of medical images is essential for reducing the cost of data storage and transmission time which in turn helps in better utilization of Bandwidth. The demand for images, video sequences and computer animation has increased drastically over the years which have also resulted in image and video compression. Image compression is broadly classified into lossy and lossless compression. Fractal image compression (FIC) is a lossy compression method. In fractal image compression an image is coded as a set of contractive transformations in a complete metric space. The set of contractive transformations is guaranteed to produce an approximation to the original image. In this paper quad -tree FIC is implemented on different Imaging modalities like Medical Resonance (MR) Image of Brain, Computerized tomography(CT) of Bone.The quality factors like Mean Square Error (MSE) , Peak Signal–to–Noise-Ratio (PSNR) Compression ratio(CR), Encoding time and decoding time for different imaging modalities with different threshold va lues are analyzed in this paper. From the matlab simulated results it is observed that Quad -tree FIC works better on medical image as it provides better PSNR, CR values over the other images. This paper also includes a comparison between standard FIC and Quad-tree FIC on MR image of Brain and study of the parameters reveals that Quad-tree FIC works better than Standard FIC. Keywords— Medical Imaging Fractal image compression, Quad-tree partitioning, objective quality measures