夜间视频增强算法研究及硬件设计

发布时间：2018-06-03 13:50

本文选题：图像处理 + 夜间视频增强　；参考：《哈尔滨工业大学》2017年硕士论文

【摘要】：视频技术是当今社会中应用最为广泛的技术之一,在安防监控、交通安全、航天航空等领域等发挥着至关重要的作用,然而在夜间环境中,由于光照不足、强度不均等原因,采集到的视频通常清晰度低、信息丢失严重,因此有必要对夜间视频进行增强处理,以提升其视觉效果和质量。基于以上背景,本文主要研究夜间视频增强算法,以及嵌入式夜间视频增强系统的设计与实现。首先,本文研究了经典的图像增强算法在夜间图像增强中的应用。针对夜间视频的特点,本文重点研究了直方图均衡、同态滤波和Retinex算法在夜间图像增强中的应用,并通过实验仿真分析了各算法的有效性和局限性。其次,本文重点研究了基于暗通道先验的去雾算法在夜间图像增强中的应用。经过实验验证,夜间图像与雾天图像的直方图分布呈现一种对称的特性,因此可以利用雾天图像增强算法实现夜间图像增强的目的。本文通过实验证明了暗通道先验算法能对夜间图像起到很好的增强效果,同时也发现该算法存在光晕伪影现象以及计算量大等不足。因此,本文提出了相应的优化方法,并通过对比实验,从主观视觉效果和客观指标两方面分析了优化算法与其他算法的性能。最后,设计并实现了夜间视频增强硬件系统。考虑到硬件系统的成本、体积和实用性等需求,本文设计了基于现场可编程门阵列(Field Programmable Gate Array,FPGA)的嵌入式夜间视频增强系统。对系统中视频处理板卡的电路设计进行了简要说明,详细介绍了基于Verilog硬件编程语言和流水线思想设计的FPGA功能逻辑模块,主要包括视频数据接收和发送、色彩空间转换、视频增强处理等功能模块,并验证了各模块的正确性。最后,在实际夜间环境中,对夜间视频增强系统的效果进行测试,实测结果表明,本文设计的嵌入式夜间视频增强系统能实现对分辨率为1920x1080、帧频为30帧每秒的视频进行实时有效的增强处理。
[Abstract]:Video technology is one of the most widely used technologies in today's society. It plays an important role in the fields of security monitoring, traffic safety, aerospace and so on. However, in the night environment, because of the lack of light, the intensity is not equal. The captured video usually has low definition and serious loss of information, so it is necessary to enhance the visual effect and quality of the night video. Based on the above background, this paper mainly studies the night video enhancement algorithm, and the design and implementation of embedded night video enhancement system. Firstly, this paper studies the application of classical image enhancement algorithm in night image enhancement. According to the characteristics of night video, this paper focuses on the application of histogram equalization, homomorphic filtering and Retinex algorithm in night image enhancement, and analyzes the validity and limitation of each algorithm through experimental simulation. Secondly, this paper focuses on the application of dark channel priori de-fogging algorithm in night image enhancement. Experimental results show that the histogram distribution of nocturnal and foggy images is symmetrical, so the enhancement algorithm of fog images can be used to achieve the purpose of nocturnal image enhancement. The experimental results show that the dark channel priori algorithm can enhance the night image very well. At the same time, it is also found that the algorithm has some shortcomings such as halo artifacts and large amount of computation. Therefore, this paper puts forward the corresponding optimization method, and analyzes the performance of the optimization algorithm and other algorithms from two aspects of subjective visual effect and objective index through comparative experiments. Finally, the hardware system of night video enhancement is designed and implemented. Considering the cost, volume and practicability of hardware system, this paper designs an embedded night video enhancement system based on Field Programmable Gate Array (Programmable Gate). The circuit design of video processing board in the system is briefly described, and the logic module of FPGA function based on Verilog hardware programming language and pipeline is introduced in detail, including video data receiving and transmitting, color space conversion. Video enhancement processing and other functional modules, and verify the correctness of each module. Finally, the effect of night video enhancement system is tested in real night environment. The experimental results show that, The embedded night video enhancement system designed in this paper can achieve real-time and effective enhancement of video with resolution of 1920x1080 and frame rate of 30 frames per second.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【相似文献】