Real-time video enhancement on FPGA by self-enhancement technique

Abstract

This paper proposes two efficient methods along with a hardware structure for real-time video enhancement, which are specific for dedicated hardware platforms - where memory's size is limited and not totally suitable with an amount of complex computation. The presented approaches here rely on Self-enhancement technique and their important coefficients are suggested through a vast range of statistics that were made by systematically analyzing from implementations using Matlab. A hardware structure is, then, designed on FPGA based on the best proposed method. In order to demonstrate the quality of the proposed, a practical environment with a FPGA board and a real camera was also set up to evaluate. Such extensive experimental results indicate that both the proposed methods and hardware structure are effective, feasible and straightforward to apply into practice.