Real-time video enhancement on FPGA by context-based fusion technique

摘要

This paper proposes a new method for video enhancement based on context-based fusion technique with respect to surveillance applications in real-time. In addition, a FPGA-based design for the proposed method is also presented. The developed technique here outperforms the original methods by without complex formulas and being time-efficient, whereas the output videos still keep a high quality. The effectiveness of this method is also demonstrated through extensive experimental results along with quantitative measures. Besides, we successfully set up a real system to assure that our hardware design is feasible and straightforward to apply into practice.