NTT DoCoMo, Inc. started an introductory service of the next generation mobile communication system, IMT-2000 from May, 2001 in Japan. Such a new service provides the image communication capability in addition with voice and data communication with a transmission rate of 384 kbit/s at maximum. This study aims to clarify a display size of a mobile-multimedia communication terminal from a viewpoint of user satisfaction. because a display size of such a terminal is restricted. We have carried out subjective evaluations on the minimum display sizes for moving images and the minimum sizes of various characters superimposed. Based on these evaluations, design guidance of the pictures is discussed.
In study of software-based video coding, there are several reports about faster video coding method achieved by efficiently omitting video coding process itself. However, the. actual speed of such a software-based video coding depends on implementation environments. Accordingly, under some environments of lower processing speed, the processor can not encode video data in real-time. Therefore we are proposing a control method of changing adaptively video coding time dependently on processing speed. In this paper, we prove the relation between image-related information calculated in SEA motion estimation method, and processing time.
The recent technological advancement has been helpful for much improvement of various environmental conditions in our society Especially, Information Technology contributes to remarkable progress in the performance of various electrical apparatus for communication and signal processing. But I think one of the most important things is not the improvement of performance but the will to design a human and heart warming device for friendly interface in our future community. From this point of view, I report on some applications of friendly interface for human communication with the aged and the people with disability.
Picture coding is divided into several components, such as motion compensation, inter/intra-mode switching, DCT, VLC etc. Each component has been improved with only the component. There arises a problem that an interaction with other components or a total optimization is not be considered. Two components at distant location are especially difficult to be treated and the relation between them is not shown. In this report, from an analysis of coded bitstream, redundancy of coded bitstream is evaluated. The analysis gives some policies to improve coding performance of each component and to clarify coding relation between components. A method of random number analysis is proposed. Entropy, Markov entropy, sampled calculation, auto-correlation are used for measuring redundancy of coded bitstream. To clarify the measured results, a precision criterion is derived, which gives the necessary data amount for valid measuring. Examples show the entropy values are at least 0.9-0.85. This results indicates that re-encoding is possible.
Distributed multimedia applications and mobile multimedia applications are developed and began to be used widely as spread of Internet and improvement of microprocessor. Such applications arerequired 1)to provide end-to-end QoS on dynamically fluctuating computation resources enviroment and 2)to meet multifarious and inconstant user's requests for multimedia services(for instance, video services and audio services). In this report, we propose Layered Adaptive QoS(LAQoS) model, which constructs multi layers based on meta level architecture in order to meet the above two requirements for distributed and mobile multimedia applications, study proposed model.
For providing distributed multimedia applications with end-to-end QoS(Quality of Service)guarantees, resource reservation based control mechanisms should be employed in both of networks and end systems. In this paper, we propose a resource allocation scheme for maximizing the user's utility, which is represented as relation between benefit obtained through allocated resources and cost paid for them. In the scheme, each client is accommodated in the proper cluster to reduce the toad of the server and the networks. We confirm that our proposed scheme can achieve an effective use of resources while providing high quality video transfer.
This report studies the effect of assignment of audio and video streams to BAs on the media synchronization quality for the AF service in DiffServ by experiment. In our experiment, we deal with two pairs of audio-video streams and an interference stream; we then measure the intra-stream and inter-stream media synchronization quality by changing assignment of their streams to BAs. The experimental results show that audio and video streams are desired to be assigned to distinct BAs from a media synchronization point of view. Furthermore, we investigate the relationship between threshold values of WRED, which is a part of the AF service architecture, and the media synchronization quality by the experiment.