Capstone Project

Back to listing
Group 2007-12 Status completed
Title Universal Video Transcoder
Supervisor Amer, A.
Description Video is becoming an integral part in various applications such as entertainment, education, medicine, databases, security, and even wireless applications. Video signals are present around us in many different types and formats. When engineering a video device for a specific application, a design decision must be made as to what types of video signals to support. However, we often require to use the device for other types of video signals which are unsupported. For that purpose, video transcoders which change the video signal from one type to another became increasingly popular and in high demand. When speaking about the type of a video signal, we mainly mean the compression technique used to reduce the size of the signal and the container or file format used to store the compressed signal into a computer file. For example, DivX and Xvid are implementations of the MPEG4 standard which defines techniques to compress the video signal and reduce its size, while AVI and MP4 (the file extensions) define ways to store the compressed signal into file. There are many cases when we need to transcode a video signal. For example, when converting a VCD movie, which is MPEG1-coded and stored in DAT file format, into resized MPEG4-coded signal that is stored in MP4 file format to play it on iPods. Another example is changing a sequence of JPEG images (MJPEG video) into an MPEG2-coded video which is stored in VOB file format to burn it into DVD. In this project, students will gain practical information and valuable insight into the different compression standards and container formats available nowadays and utilize this information in developing a universal video transcoder. The universal transcoder will work in two phases. In the first phase, a decoder will be used to convert the compressed video signal into uncompressed raw data. In the second phase, an encoder will recompress the raw data into the desired format.
Requirement Students should have a strong knowledge in object oriented C++ and the Linux operating system. Background in Signal Processing and/or Pattern recognition is an asset.
Tools Linux
Number of Students 3
Students Jascha Jacques Qi Li Wei Yong Ng
Comments:
Links: