基于嵌入式系统的透雾监控系统的设计与实现

发布时间：2018-04-28 11:34

本文选题：视频监控 + S3C2440A　；参考：《太原理工大学》2017年硕士论文

【摘要】：视频监控技术伴随着网络、计算机、芯片、存储、云计算、物联网、大数据等技术的发展而迅速发展,在边防、安防、交通、金融等领域都有广泛的应用,但是目前常见的雾霾天气,使得获取的视频存在颜色失真、模糊不清、对比度降低等问题,对室外监控的视频信号采集及视频信息提取产生了重大的影响。因此,改善雾霾天气下图像的质量,提高监控系统的透雾性能显得尤为迫切。本文首先介绍基于物理模型与非物理模型的去雾算法,分析优劣,将算法组合优化,得到去雾效果良好的算法;然后搭建基于S3C2440A与TMS320DM648的嵌入式硬件平台,S3C2440A为主控单元,DM648为算法运行单元;最后针对嵌入式平台以及移植过程对算法进行再次优化,采用算法透雾的方式实现雾霾天气图像视频的清晰采集,主要研究内容如下:1、基于大气散射理论,研究雾霾天气下图像质量降低的原因,研究基于物理模型的暗原色先验算法与基于非物理模型的多尺度Retinex(MSR)算法,并用MATLAB仿真,分析各算法优缺点,提出改进算法并对改进算法进行仿真验证。2、基于视频采集、处理、输出的研究目的,搭建基于S3C2440A与DM648的嵌入式硬件平台,搭配摄像头作为核心硬件开发平台与存储、网络传输等芯片共同实现研究目的。3、编写核心处理程序。核心处理程序主要包含操作系统程序、摄像头驱动程序,视频采集程序,透雾算法程序。针对硬件平台特性对算法程序做实时性优化,将调试好的程序移植到硬件平台上,使搭建的硬件平台在雾霾干扰条件下能够获得比较清晰的图像。通过理论分析以及对硬件平台、去雾算法程序的多次实践、仿真验证,得出本文所做研究能够较好的获取、恢复出清晰的图像,图像质量有明显提高。
[Abstract]:With the rapid development of network, computer, chip, storage, cloud computing, Internet of things, big data and other technologies, video surveillance technology has been widely used in border defense, security, transportation, finance and other fields. However, the common haze weather makes the obtained video have color distortion, ambiguity, low contrast and other problems, which has a significant impact on outdoor video signal acquisition and video information extraction. Therefore, it is urgent to improve the image quality in haze weather and improve the fogging performance of the monitoring system. Firstly, this paper introduces the de-fogging algorithm based on physical model and non-physical model, analyzes the merits and demerits, and optimizes the algorithm to obtain a good de-fogging algorithm. Then build the embedded hardware platform based on S3C2440A and TMS320DM648, S3C2440A as the main control unit and DM648 as the algorithm running unit; finally, the algorithm is optimized again for the embedded platform and the transplant process. In this paper, the clear acquisition of haze weather image and video is realized by using the algorithm through fog. The main research contents are as follows: 1. Based on the theory of atmospheric scattering, the reason of image quality degradation in haze weather is studied. In this paper, the dark priori algorithm based on physical model and the multi-scale Retinexan MSRs algorithm based on non-physical model are studied. The advantages and disadvantages of each algorithm are analyzed by MATLAB simulation, and the improved algorithm is proposed and simulated to verify the improved algorithm. The algorithm is based on video capture and processing. The research purpose of the output is to build the embedded hardware platform based on S3C2440A and DM648, and combine the camera as the core hardware development platform and storage, network transmission and other chips to achieve the research goal. 3, write the core processing program. The core processing program mainly includes operating system program, camera driver program, video capture program, fog-permeating algorithm program. Aiming at the characteristic of hardware platform, the algorithm program is optimized in real time, and the debugged program is transplanted to the hardware platform, so that the hardware platform can get clear image under the condition of haze interference. Through the theoretical analysis and the practice of the hardware platform and the de-fogging algorithm program, the simulation results show that the research in this paper can obtain the clear image and improve the image quality.
【学位授予单位】：太原理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TN948.6

【相似文献】