Design of High-efficiency Natural Image Embedded Hardware Surveillance System

Although the surveillance system can capture images loyal to the realistic scene, the captured images may not have clear texture performances, and the characteristics of brightness cannot provide the Human Visual System (HVS) pleasure sense. In view of this, many surveillance systems in the market have also added contrast enhancement functions on line with market demand. Although the current surveillance system products on the market have a contrast-enhancement architecture, they sacrifice the necessary real-time processing performance of the surveillance system, resulting in extremely large memory usage and long computing time. Therefore, this work aims to improve Parametric-Oriented Histogram Equalization (POHE), design an improved integral image algorithm in the hardware architecture, and solves the problem of excessive memory usage of the original POHE by using a counter.