FPGAs show interesting properties for real-time implementation of video processing systems. An important feature is the available on-chip RAM blocks embedded on the FPGAs. This paper presents an analysis of the current and future requirements of video processing systems put on these embedded memory resources. The analysis is performed such that a set of video processing systems are allocated onto different existing and extrapolated FPGA architectures. The analysis shows that FPGAs should support multiple memory sizes to take full advantage of the architecture. These results are valuable for both designers of systems and for planning the development of new FPGA architectures

This paper presents an algorithm for the allocation of on-chip FPGA Block RAMs for the implementation of Real-Time Video Processing Systems. The effectiveness of the algorithm is shown through the implementation of realistic image processing systems. The algorithm, which is based on a heuristic, seeks the most cost-effective way of allocating memory objects to the FPGA Block RAMs. The experimental results obtained, show that this algorithm generates results which are close to the theoretical optimum for most design cases.

FPGA offers the potential of being a reliable, and high-performance reconfigurable platform for the implementation of real-time video processing systems. To utilize the full processing power of FPGA for video processing applications, optimization of memory accesses and the implementation of memory architecture are important issues. This paper presents two approaches, base pointer approach and distributed pointer approach, to implement accesses to on-chip FPGA Block RAMs. A comparison of the experimental results obtained using the two approaches on realistic image processing systems design cases is presented. The results show that compared to the base pointer approach the distributed pointer approach increases the potential processing power of FPGA, as a reconfigurable platform for video processing systems.

This paper presents a tool for automatic generation of the memory management implementation for spatial and temporal real-time video processing systems targeting field programmable gate arrays (FPGAs). The generator creates all the necessary memory and control functionality for a functional spatio-temporal video processing system. The required memory architecture is automatically optimized and mapped to the FPGAs' memory resources thus producing an efficient implementation in terms of used internal resources. The results in this paper show that the tool is able to efficiently and automatically generate all required memory management modules for both spatial and temporal real-time video processing systems.