Research Projects

	Strategy

Video Coding

Motivation: Efficient video coding is an enabling technology for many applications today and in the future.  Reducing the bit rate that is needed to represent a video signal in a service to an amount that makes the service economically viable will remain one of the highest priorities in the development of these services. Therefore, the Image Communication Group focuses on video compression and has so far been actively contributing to this subject. Links to selected project descriptions:

• H.264/MPEG4-AVC
• Context-Based Adaptive Binary Arithmetic Coding (CABAC)
• Rate-Constrained Coder Control
• Multi-Frame Motion-Compensated Prediction
• SVC: Scalable Extension of H.264/AVC
• Video Coding Using High-Resolution Mosaics

Multimedia Transmission

Motivation: The efficient transmission of multimedia is a key component in current and future services. The work of the Image Communication Group in this area focuses on the interaction between video coding and the transmission characteristics of wired and wireless networks including used protocols and error protection mechanisms. For the verification of the proposed methods and for feasibility studies, multimedia transmission systems are implemented.  Links to selected project descriptions:

• Joint Source/Channel Coding
• Adaptive Real-Time RTP Streaming and IPTV over the wired Internet
• Multimedia Transmission over 3G Mobile Networks
• TV over DVB-T Using H.264/MPEG4-AVC
• Mobile TV: DVB-H and DMB
• Layer-Aware Forward Error Correction (L-FEC)
  

Semantic Approaches in Image Processing

Motivation: The increasing amount of produced and consumed digital picture and video data makes the utilization of semantic approaches in image processing more and more important. Methods for extraction and deployment of semantic information are still in their infancy and are being used in research by the Image Communication Group for image search in databases and for improved video coding. Links to selected project descriptions:

• Search Engine for Pictures
• Video Coding using Semantic Texture Analysis and Synthesis
• 3D Model-Aided Video Coding

Coding, Interactive Streaming, and Visualization of 3D Environments

Motivation: Novel technologies for coding, transmission and visualization of 3D environments enable a number of applications such as free viewpoint video or virtual guides  (landscapes,cities,buildings). The steady growth of computational power, memory and transmission capacity will enable such applications to become economically viable in the near future. Links to selected project descriptions:

• Multi-view Video Coding
• Time-varying 3D Computer Graphic Models and Free Viewpoint Video
• Augmented Reality
• Interactive Streaming of a 360° Panoramas
• Optical Traffic Surveillance