Video Dehazing Hardware Accelerator Design based on Dark Channel Prior with Sky Preservation

• Published in: Proceedings of Technical Program of International Symposium on VLSI Design, Automation and Test pp. 1 - 4
• Main Authors: ZHAO, ZI - YI, XIAO, AN - TAI, GUO, JIUN - IN
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.08.2020
• Subjects: Atmospheric modeling; Estimation; Filtering algorithms; Hardware; Random access memory; Reduced instruction set computing; Semiconductor device modeling
• ISSN: 2472-9124
• DOI: 10.1109/VLSI-DAT49148.2020.9196353

With the increasingly serious haze disaster, the development of dehazing technology is highly concerned. But the current software dehazing algorithm suffers from too high computational complexity. We selects the algorithm based on dark channel prior with sky preservation [1] as the base software algorithm, and a hardware accelerator is designed to simplify the data transmission process and optimize the data processing. The whole operation simulation of the accelerator is realized in NC-Verilog hardware design environment. The accelerator can achieve 75fps@640x480 video dehazing under 0.18\mum CMOS technology. It can be applied to driving assistance and surveillance system.