Detta är ett uppsatsförslag hämtat från Nationella Exjobb-poolen. Klicka här för att komma tillbaka till samtliga exjobbsförslag.
Video decoder concealment
Modern video compression algorithms utilize both temporal and spatial correlation. For example, when encoding a video frame, it is very common that the previous frames are used as a prediction base. The difference between the prediction from the previous frame(s) and the current frame is then encoded.
Modern video standards are becoming better at prediction which results in greater compression. This large use of prediction does however result in significant quality loss when frames are lost. This can also result in the propagation of errors through a video sequence.
Error concealment is a way for a decoder to “cover up” these errors by substituting the missing pieces of video, with something either based on older “correct” frames, or other parts of the same frame.
This thesis involves studying and implementing existing concealment algorithms as well as developing a new concealment algorithm.
The objective of this thesis work is to study and optimize video decoding in lossy networks by concealing errors as much as possible. Both low-bitrate mobile phone video as well as digital TV video will be studied. The main work will be done using existing H.264 video encoder/decoder software (C code) developed at Ericsson Research.
The work will contain:
• Theoretical studies
• Implementation of existing algorithms
• Developing a new concealment algorithm
• Analysis of decoder performance
• Technical report
We are looking for a bright M.Sc student with interest and knowledge in algorithm design. Good programming skills in C or the ability to learn it fast is also a requirement.
Ericsson Research – Multimedia technologies
The thesis work will be done at the Visual Technology section of Multimedia Technologies at Ericsson Research in Kista, Torshamnsgatan 23.
Clinton Priddle, [email protected] , 08-757 3143
Applications (containing a short personal letter, recent grades and CV) should be sent by e-mail to: [email protected]
Please apply before 2007-08-15
Informationen om uppsatsförslag är hämtad från Nationella Exjobb-poolen.