Research On Real-Time SoC For Video Image Processing

Video & image processing is widely used in those fields: digital TV, videoconferencing, digital libraries, distant learning, telemedicine, interactive games. And it places very high demandson devices for transmission, storage, and computation.The algorithms of video & image processing are anysised in this thesis. The video processing tasks are classified into three levels according to the scale of computation, parallelity, complexity, and regularity. On the base of the classfication, these tasks are mapped into hardware or software with appropriate architecture.Video & image processing system is a typical real time system. In this paper, it抯 focused on the time requirement for real time system, especially for the HDTV video decoding system because it is regarded as the highest time request which is as six times as SDTV. Then it抯 representative.There are two kinds of architecture for video & image processing system: programed and dedicated. The former is more flexible and esay to update. But the hardware is complicated, and power consumption is high. The later is designed and optimized against a set of algorithms. It saves more hardware and has higher processing speed. With the development of VLSI technology, the trend.is to develop System on a Chip (S0C). We focus on dedicated SoCBy now, the speed of logic computation in processing units is very high with general CMOS technics. However, the speed of the storage memory and the data transform between memory and processing units is not as high as the processing units. It becomes the bottlenneck of the system. With ihe evolution of the video & image compressing algorithms, the structure of data is more complex and more correlative with each other. In this paper, data storage structure and data bus schedule in SoC are researched to resolved those prominent problems:Firstly, two kinds of data storage structure are studied: local memory and frame memory. The structure in the future is forwarded.Secondly, the bus data schedule is studied based on the sotrage structure: time division multiplexed, round robing, fixed priority schemes. They are adopted in different application fields because of their features. In video & image processing, the data are continous, periodical, correlative. According to the characters, a hybrid scheme is developed.Finally, as an example, HDTV video decoding system verifies those proposed schemes of data storage and bus schedule. It meets the requirement of decoding in real time, and it reduces the cost of hardware and power.