Research On Haze Image Detection And Video Dehazing Algorithms Based On FPGA

In the foggy weather,the captured outdoor video images showed poor picture quality,weak clarify and low contrast.As an important member of artificial intelligence,machine vision is developing rapidly,and the preconditions for artificial intelligence systems to obtain target features through machine vision are often based on clear video images.Therefore,the research of haze image detection and video defogging algorithm in outdoor environment is particularly important.With the rapid development of microelectronics technology and semiconductor technology,the research on haze image detection and video defogging algorithms is not limited to PC hardware platforms.This article focuses on the characteristics of FPGA chips and detects haze images based on dark channel prior theory.Optimized and improved the video defogging algorithm and designed a set of FPGA-based haze image detection and video defogging system.The work of this paper is as follows:(1)Research on FPGA-based fog and haze image detection algorithm.Based on the concept of the dark channel prior theory and the phenomena discovered by the theory,complete the quantitative evaluation and mathematical analysis of the characteristics of the haze image,and finish the automatic detection of the haze image by the system.(2)Research on dehazing algorithm based on FPGA.Analyze the atmospheric physical model and confirm the unsolved parameters in the process of image defogging.From the time and space complexity of the algorithm,using hardware description language compilation,the implementation method and computational complexity of each part of the algorithm are analyzed,and the defogging algorithm is improved and optimized to make it suitable for FPGA hardware implementation.(3)Hardware platform construction.Hardware platform with FPGA as the main control chip was built to provide a hardware foundation for real-time video acquisition,real-time defogging processing,and real-time video display functions.(4)Logic circuit design.Use Vivado HLS synthesis tool and Verilog HDL language to obtain the transplantation of haze image detection and video defogging algorithms,and complete the logic circuit design of video acquisition module,algorithm module,data buffer module and video display module.Through algorithm simulation and hardware test results,it was proved that the smog image detection algorithm in this paper can finish the automatic detection of smog image.Compared with the original algorithm,defogging effect is slightly worse,but the improvement in algorithm efficiency is considerable.The FPGA-based haze image detection and video defogging system designed in this paper can not only achieve video defogging processing,but also meet real-time requirements.