Source coding of speech and video signals
Article Abstract:
The 1980s saw rapid progress in coding techniques for speech and video. Digital techniques are always used in source coders for speech signals. Speech is treated as a single full-band signal by time-domain coders. Predictive coders use linear prediction to remove the time redundancy in the signal. The primary difference in the different algorithms is in the degree of prediction and the quantization strategies. The main task of the bit-rate reduction process is removing spatial redundancy. New algorithmic developments are made possible by progress in VLSI. Emerging techniques, such as multirate signal representations and motion-compensated interpolation, may unify all video compression techniques in a framework of hierarchical coding algorithms. Currently, combinations of DCT and sub-band coding produce the best results.
Publication Name: Proceedings of the IEEE
Subject: Electronics
ISSN: 0018-9219
Year: 1990
User Contributions:
Comment about this article or add new information about this topic:
Satellite broadcasting of time and frequency signals
Article Abstract:
International standards for the accurate and precise dissemination of time and frequency information have assumed great importance, due to the demands of such fields as satellite communications, geophysics, navigation and radioastronomy. Most dissemination techniques are based on radio transmission, either through dedicated broadcast or through signals intended for established systems such as television or navigation. The advantages of satellite-based time broadcasting over traditional ground-based methods are presented. Several one-way satellite time services are discussed, particularly a service using the Indian INSAT domestic satellite. Its signal format, transmission and receiving setups and satellite position prediction methods are described in detail.
Publication Name: Proceedings of the IEEE
Subject: Electronics
ISSN: 0018-9219
Year: 1991
User Contributions:
Comment about this article or add new information about this topic:
Reconstruction of two-dimensional signals from level crossings
Article Abstract:
Reconstructing two-dimensional signals from partial information plus constraints is an on-going area of research. Two sampling strategies are used in this research: explicit, which is based on prespecified sampling constants, such as Nyquist sampling, and implicit, where the signal assumes prespecified values. Semi-implicit and implicit sampling strategies can be used to reconstruct signals from multiple-level threshold crossings. There are two reconstruction algorithms: a linear least-squares approach and an iterative algorithm based on the theory of projection on convex sets. Reconstructing from multiple-level threshold crossings fills in the gap between explicit, semi-implicit, and implicit sampling strategies.
Publication Name: Proceedings of the IEEE
Subject: Electronics
ISSN: 0018-9219
Year: 1990
User Contributions:
Comment about this article or add new information about this topic:
- Abstracts: The control of discrete event systems. Algebras of discrete event models
- Abstracts: Land mobile satellite systems. Domestic and regional satellite systems
- Abstracts: The electrical properties and device applications of homoepitaxial and polycrystalline diamond films
- Abstracts: Nonintrusive appliance load monitoring. Short-term load forecasting
- Abstracts: Petri nets: properties, analysis and applications. Entropy nets: from decision trees to neural networks. Modern permanent magnets for applications in electro-technology