NHK Science and Technical Research Laboratories are studying new data broadcasting services using the 12 GHz band satellite television broadcasting system. In this paper, an enhanced teletext service called PRESENT(Personal REquest Service by ENhanced Teletext) is proposed. The features of this service are as follows: - Each program is composed of multimedia such as text,image and still picture. - Each program is structured from both the layout structure and the logical structure by the rules of ODA. - Index data assist the users in selecting the programs which they want to receive.

View full abstract

The enhanced teletext named PRESENT is a new data broadcasting service which is transmitted by BS data channel and ISDB. In this service programs consist of multi-media data, index-data and structure-data. Index-data are used for program selection. And structure-data are used for making display-format flexible. Three experimental receivers were developed for testing these functions , and evaluated human-interface and retrieval performance.

View full abstract

We have developed a simplified MUSE decoder which is built in an HDTV. We have simplified stationary image processing parts which a full-spec. MUSE decoder normally has, so that the received MUSE signal is always processed as a moving image in this decoder. Aliasing distortion is caused in such a way that temporally sub-sampled MUSE signal is interpolated spatially There are two kinds of aliasing distortion caused in this case. We succeed in removing both of these aliasing distortions using temporal filtering processing which is suitable for each aliasing characteristic. This paper mainly shows the way to reduce these aliasing distortions in this decoder.

View full abstract

We have developed LSIs for wide screen TV reciever. These LSIs have functions such as vertical expansion and horizontal compression for interlaced scanning type reciever and progressive scanning type reciever, using digital signal processing. This paper shows the functions, vertical expansion method, and applications.

View full abstract

This paper describes hardware implementation for letter-box EDTV decoder, The authors have examined the decoder in which frame memory is commonly used for main panel signals and bar signals. As a result, the examination has proven that the decoder needs to add about 20 % more memory to the EDTV-1.

View full abstract

This paper studies a VT (Vertical-Temporal) helper signal on EDTV. The signal processing of an LD (Line Difference) signal used for the VT is analyzed in the vertical-temporal frequency domain. The LD processing makes psychovisual S/N degraded. Therefore, non-linear amplitude expansion and compression technique is adopted for noise reduction, but it causes a flicker. A new motion-adaptive LD signal processing at receivers is studied to improve the flicker.

View full abstract

This Paper presents a study of a three-dimensional spectrum of a transmitted noise on a vertical-Temporal (VT) helper signal in wide-aspect EDTV, and a proposal for a new noise reduction scheme. In this scheme, a vertical-temporal low frequency component of the signal which dominantly causes noise is reduced at the receiver without degrading the reproduced Picture quality. A Simulation experiment shows the effect of the noise reduction.

View full abstract

A system configuration of EDTV II audio which can provide high quality PCM sound service on terrestrial broadcasting is described. Original PCM sound data is compressed to about 1/6 using source coding, such as MPEG sub-band coding algorithm. The compressed data is arranged with error correction code and data spectrum is shaped by partial response coding. The data is transmitted via QAM (quadrature amplitude modulation) of a video carrier, that has data transmission bit rate from 300kbps to 1200kbps. We evaluated bit error rate of transmissions degraded by linear distortion, such as ghost, group delay distortion and demodulation phase error. The experimental results proved BER can be improved to 10^<-5>-10^<-7> by waveform equalization with 6dB D/U ghost or 30dB video S/N condition.

View full abstract