EDGE PRESERVING IMAGE COMPRESSION FOR MAGNETIC RESONANCE IMAGES USING DANN-BASED NEURAL NETWORKS

EDGE PRESERVING IMAGE COMPRESSION FOR MAGNETIC RESONANCE IMAGES USING DANN-BASED NEURAL NETWORKS

ABSTRACT

This seminar report gives us an overview of the image compression of Magnetic Resonance Images. It gives us the steps involved in image compression and the characteristics of an appropriate image compression scheme. It covers the DCT Technique and Vector Quantization .

The seminar report covers the data compression using Neural Networks and gives the architecture of Edge Preserving Image Compressor.
It gives the compressor and de-compressor overview. The report also covers the goals for developing a compression architecture.

The report addresses the use of DANN-Network for the compression of Magnetic Resonance Images. It covers the features for improving EPIC upon the DANN architecture. The report gives details about the Modified DANN Training procedure. It also covers the Epic Compression Algorithms.

Finally the report gives the advantages of Epic Compression scheme and also the future enhancements.