• MPEG-4 Video Codec
• H.263 Video Codec

MPEG-4 Video Codec

A Schematic Overview of MPEG-4 Video Codec
The MPEG-4 system includes the presentation and transportation of digital audio and video. The concept of the MPEG-4 Codec is described as the following functionality block diagram:

Pre-processing
For reaching the best compression capability, the pre-processing block performs all necessary processing on the input video from the camera till it reaches the suitable color format.
The functions of pre-processing include:

• Color space transformation: RGB to YUV.
• Noise reduction: the improvement of picture smooth without affecting sharpness.
• Scaling of the input video from the CMOS sensor resolution to the standard QCIF or CIF sizes.
• Stabilization: canceling of camera movements.

Video Encode
The encoder of Brogent MPEG-4 Codec captures and encodes the video to the MPEG-4 bit-stream showed as the following block diagram:

Video Decode
The MPEG-4 bit stream is initially uncompressed from the bit stream using Huffman tables to produce a two dimensional array of coefficients.
The two-dimensional array of coefficients is dequantised to produce the reconstructed DCT coefficients. This process is essentially a multiplication by the quantiser step size. The quantiser step size is calculated by two mechanisms; a weighting matrix is used to modify the step size within a block and scaling factor. After the dequantisation arithmetic the resulting coefficients are saturated and mismatch control operation is performed to give the final reconstructed DCT coefficients.
The iDCT is then performed. Motion compensation determines the motion vectors that are used for the basis of the predicted blocks for the current image, by copying blocks from the previous image. The motion vectors are extracted from the bitstream, and the predicted blocks are taken from the previous image that is stored in the frame buffer. The resultants from the motion compensation and iDCT algorithms are added together to from the YUV-image. This image can then be post-processed prior to display.

Post-processing
The post-processing block performs all necessary processing on the video which is output from the decoder, until it reaches the required format for display.
The functions of post-processing include:

• De-blocking and de-ringing algorithms.
• Scaling of the output video from QCIF or CIF to the required resolution of the screen.
• Color space transformation: converting to fit the specific display.

Appendix
MPEG-4 Simple Profile
An excellent general purpose video codec, with efficient, error resilient coding of rectangular video objects, ideal for applications of surveillance, network cameras, video conferencing, multimedia gateways, head end multimedia systems, video servers and consumer electronics.

Table: Requirements of MPEG-4 simple profile.