FPGA based dynamic reconfigurable implementation of image processing algorithms

摘要

This paper presents a study on FPGA based dynamic reconfigurable implementation of image processing algorithms. Nowadays FPGAs are being widely used in hardware accelerator implementations. The most important advantage of the FPGA devices is their flexibility that comes up from their programmability feature. In a dynamic reconfigurable implementation, the programmability characteristic takes a more important role and helps in more powerful design and development of a system. The most important feature of reconfigurable systems is that they can change their resources without affecting other parts of the system. Therefore image processing algorithms and systems are more suitable for reconfigurable implementation. Face detection, fingerprint detection, and object tracking are some of the most important challenges in image processing that are being improved with partial reconfigurable implementation. The results obtained from this study shows that the algorithms that apply different simple procedure on a large number of data has a higher gain when implemented reconfigurable.