MOTION COMPENSATOR DESIGN AND INTEGRATION DECODER H.264
Digital video transmission technology has been commonly used these days. In this thesis, the author designs Motion Compensator and the Integration of Decoder System based on MPEG4 H.264/AVC. Decoder has a function to reconstruct image from the compressed original image. Three main components of M...
Saved in:
| Main Author: | |
|---|---|
| Format: | Theses |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/77249 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Digital video transmission technology has been commonly used these days. In this
thesis, the author designs Motion Compensator and the Integration of Decoder
System based on MPEG4 H.264/AVC. Decoder has a function to reconstruct
image from the compressed original image. Three main components of MPEG4
H.264/AVC decoder are Inverse Block Transform, Deblocking Filter, and Motion
Compensator. Motion Compensator produces prediction frame from one or two
references that have been chosen from some reference pictures. Inverse Block
Transform does decompression process to the input that has been accepted from
NAL (Network Abstraction Layer) and entropi decode. Deblocking Filter reduces
distortion in inter block border in a macro block.
Motion Compensator is designed to support the operation with half pixel and
quarter pixel resolution. Integration process requires addressing module and
efficient buffer design. Addressing module eases data replacement and data’s
function and setting identification in a frame. Buffer adjusts data flow inter Decoder
modules. The design is done in RTL using HDL Verilog language.
The testing is done by comparing file which is the design result with reference
data that has been obtained from reference software H.264 jm11.0 from standard
ITU-T. Then, it is continued by implementation process using Synopsys Design
Analyzer®. Beside for the implementation in standard cell, this process also
shows that the design result can be synthesized in logic level. The result obtained
displays pixel difference between simulation data and reference data is
approximately 5%. This difference is still tolerated because it can be considered
as lossy compression and does not influence picture quality seen from the sight
aspect. Time needed for decoding 30 frames is 4.936.140 clk. It shows that the
system can run real-time with minimum speed 50MHz. |
|---|