Video transmission to mobile phone

• Main Author: Ong, Magdalene Hui Shan.
• Other Authors: Lau Chiew Tong
• Format: Final Year Project
• Language: English
• Published: 2009
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: http://hdl.handle.net/10356/16927
• Institution: Nanyang Technological University

In the context of today’s multimedia environment, there has been a growing need for multimedia service, wireless channels are characterized by varying bandwidth and bursty errors, compressed video is very sensitive to such error-prone transmission channels, video coding for mobile wireless transmission becomes a challenging task. Portable, battery-operated mobile devices are faced with constraints like battery life, memory size or bandwidth of communication link. Different video compression standards are specifically designed to cater for different individual applications. There is no one universal coding technique that can be applied to different kinds of applications. A possible means can thus be adapting the best available coding strategy to different portions of the data dynamically. Dynamic video coding is a combination of multiple compression techniques dynamically adopted according to the conditions defined by the application. The video data is first divided into several coding units, depending on the level of granularity one wishes to attain. This scheme is thus to choose the best coder on each video sequence under a global rate constraint, with regards to its scene statistics and content. It comprises of optimizing a global rate distortion with the selection of a coder and its parameters on each video sequence. This is established by adopting the Lagrangian optimization with the selection of a Lagrange multiplier, λ obtained through an algorithm utilizing bisection search which will be elaborated in detail in this report. Comparisons were made with screenshots of YUV sequences being encoded with different state-of-the-art coders. Rate-distortion optimization takes place after the generation of rate distortion curves for each video sequence.