TSBK35 · 6 credits · Advanced level (A1X) · Datateknik, Elektroteknik, Medieteknik
Audio and Image Compression
Discontinued · replaced by TSBK38
No exam statistics yet
There is no exam sitting with at least 5 students in the last three academic years. New courses and project-only courses often have no statistics.
Part of 2 programmes
About the course
Aim
The course should give knowledge about methods used for data compression and how these methods are applied to audio and image signals. After taking the course, the student is expected to be able to
- Make random models for sources
- Analyze random source models
- Understand source coding methods such as huffman coding and arithmetic coding.
- Understand quantization
- Understand linear predictive coding
- Understand transform coding.
- Understand subband coding.
- Calculate coding performance for different coding methods, given random models for the sources
- Know how different coding methods are used for compression of still images, video and audio data.
- Know what methods are used in different compression standards (GIF, PNG, JPEG, MPEG, mp3, AAC, speech kodning).
Content
Theory:
- Random signal models
- Lossless coding
- Entropy
- Huffman coding
- Arithmetic coding
- Lempel-Ziv coding
- Scalar quantization
- Vector quantization
- Linear predictive coding
- Transform coding
- Subband coding.
- Motion estimation and motion compensation.
- Still image coding: GIF, PNG, JPEG, JPEG-2000
- Video coding: MPEG-1/2/4
- Speech coding: CELP, GSM
- Audio coding: MPEG-1 audio layer I/II/III, Dolby Digital, AAC, Ogg Vorbis
- Examiner
- Harald Nautsch
- Scheduled hours
- 46 h
- Main field of study
- Datateknik, Elektroteknik, Medieteknik