Image Compression and Packet Video
525.759

---

Project Guidelines and Topics

Fall Semester, 2009

Project Content:

A project usually consists of reading of papers in a chosen area, compare performances of different approaches, and perhaps propose your own algorithms/improvements, implement one or two of them. Alternatively, it can be a survey of system development and  commercial deployment. The project may also be an in-depth study of a video compression algorithm or component.

All project submittals will be in softcopy form, and posted on the class web page. Submittals will be in Microsoft Word format or HTML. Supporting software and relevant reference papers in PDF format will also be submitted. 

Project Teaming:

Software oriented projects, and projects that have a large scope (more work than one person) may be submitted by teams of no more than two people (Subject to approval by Dr. Beser).

Deadlines:

• Submit a project proposal (describing the problems to be explored, include a list of references already read, and what you plan to accomplish): 10/13/08. (You can discuss with me in the next four weeks about the possible topics, either by phone or through email).
• Status Report - Preliminary findings, changes in direction, reduction in scope, etc. 11/10/08.
• Final project presentation (a 15 minutes oral presentation, possibly with demonstration of computer implementations): 12/08/08.
• Final project report: 12/08/08.

Suggested Project Topics:

• Video sampling and sampling rate conversion

• Analysis of interlaced video, deinterlacing, etc. [Beuker94]
• Conversion between chosen formats (with computer implementation) . For example, 10 f/s to 20 f/s, 24 f/s to 30f/s (film to TV), and 50 f/s to 60 f/s (PAL to NTSC).
• NTSC RF signal to RGB Conversion in software (Sampled RF data is available)

• Motion estimation

• Survey and comparison of block based techniques, fast algorithms
• Advanced algorithms (global motion estimation, mesh-based, object-based, dense motion estimation, etc.)
• 3D motion estimation

• Video compression

• Compare different coding standards (MPEG1,2,4,H.261/H.263), their performances and intended applications, availability and applicability of commercial products
• Embedded/scalable video coding
• Rate control in video
• Video transcoding in compressed domain

• Error control in video transport

• Error resilient video coding
• Error concealment techniques
• Internet Protocols enabling real-time applications (RTP/RSVP/RTSP, etc.)
• Video over wireless
• Video over Internet
• Commercial products for interactive and streaming applications

• Scene segmentation

• Region segmentation
• Object segmentation (ex.  separate the people in a scene or in a room, segment the moving foreground from stationary background).
• Object tracking

• Video analysis and summarization

• Video shot/scene segmentation
• Key frame selection
• Video retrieval based on motion/color similarity
• Video mosaicing (putting multiple videos taken from different views into a panoramic of the scene.)

• Visual/multimedia communication systems

• Digital TV / HDTV deployment in US/Europe
• MPEG-4 terminals
• Video-on-demand

• Video watermarking
• Video processing

• Deblurring and denoising with motion compensationed filtering
• Video stabilization (stabilize video taken from a shaky camera!)
• Video Exploitation Methods Survey

• Forensic Video Methods

• Enhancement Methods
• Subject and Scene Analysis Methods
• Security Video Demultiplexing (Camera ID)
• Video Authentication (Sample video from Hollywood TV and Movie can be processed)
• Video Authentication Forensics Case (Pending Approval from Subject)
• Video Authentication and the effect of video presentation format on detectability

• Video Content Searching

• SMPTE Metadata Standard (Used for NIMA UAV Engineering Data)
• Video Content Search Systems
• Video Content Search Algorithms

• Future Video Compression Standards

• MPEG-4 Part 10 as a replacement to MPEG-2

• Video Compression for Security Systems
• Tactical Video Compression Systems

Note: Part of this project list was derived from Yao Wang's Class Web Page.