Shashank's reports log

A software simulation of MPEG encoder/decoder has been discussed in this report. I have made an attempt to describe the functions of the more important blocks, along with their implementation in C and their respective simulation results.I have not discussed the theory of MPEG in detail, as I expect the reader to have a working knowledge of MPEG 2 encoder and the decoder. Please refer to the appropriate references as given at the end of the report to get a good knowledge of MPEG2. This MPEG encoder/decoder provides a compression ratio in excess of 1:10. Currently, it accepts only monochrome images. So there is no mention of the popular 4:2:0 or 4:2:2 image formats.

Motion compensation is also implemented, though to a limited extent. But the basic principle and the code can be used to implement more complex algorithms. The same is true for Huffman coding.

The size of the image used is 64 x 64 pixels, the microblocks are of size 8 x 8 and the macroblocks are of size 16 x 16. Any change to this will require the user to change the corresponding values in the source code of the encoder and decoder. In a nutshell, this is a very basic implementation of the Mpeg 2 encoder and decoder.

The entire program is coded in C, using Microsoft VisualC++ (Ver 4.0) and Borland C++ (Ver. 5.01). Only the later software is available in the EECS labs. This software is primarily designed for PC?s, running windows ?98. However, I think that it should also work on UNIX.

:: [reports] Mpeg2 encoder/decoder. ::
HOME

[reports] Mpeg2 encoder/decoder.