基于暗原色的雾天图像增强算法
时间:2018-08-28 21:17 来源:毕业论文 作者:毕业论文 点击:次
| 摘要雾天条件下，由于户外场景受到雾气的影响，导致传感器获取的图像对比度下降、图像场景细节损失，从而使得图像变得模糊，给后续的图像处理和分析带来极大的困难。所以研究雾天条件下的图像复原技术可以增加图像的对比度和清晰度，改善图像的质量，这对于保证计算机视觉系统能够在恶劣天气下可靠、稳定的工作，具有十分重要的理论和应用价值。目前对雾天图像的复原技术研究，主要有两个方向：一是从图像增强的角度进行处理，另一种则基于雾天的物理成像模型对雾天图像进行处理。图像增强的方法尽管能在一定程度上改善视觉效果，但通常情况下并不能有效去除雾的影响。相较而言，基于成像模型的方法由于具有明确的物理解释，具有更好的效果和普适性。 目前基于物理成像模型的图像去雾方法的研究也比较多，本文基于 C语言根据大气散射模型和暗原色先验理论，实现对于雾天彩色图像的去雾处理的工作较为有效。能够去除雾霾的影响，提升图像的对比度，增加能见度,实现雾天图像的快速、有效处理。 27583 毕业论文关键字 暗原色先验 图像去雾 C语言 大气散射模型 Title Single Image Haze Removal Using Dark Channel Prior Abstract Under foggy conditions, due to outdoor scenes affected by fog, the image contrast of image from sensor drop and have loss of detail in the image scene, so the image becomes blurred, and the subsequent image processing and analysis has brought great difficulties. Therefore, image restoration technology research under foggy conditions can increase the image contrast and sharpness, improve the quality of the image, which for computer vision system to ensure reliable and stable work in inclement weather, with a theoretical and practical value is very important. Currently, there are two main directions for recovery technology research of foggy image. One direction is the processing from the perspective of image enhancement, and another physical imaging model is based on a foggy image processing. Although Image enhancement method can improve the visual effect to some extent, but usually can not effectively remove the impact of fog. In contrast, the method of the imaging model-based has the clear physical interpretation, with better results and universality. Currently, the studies of image defogging physical imaging model-based approach will be more and more. The paper is based on C language , atmospheric scattering model and a priori theory of dark colors to achieve the work to fog foggy color image processing more effective. Remove the effects of haze, enhance image contrast, increase visibility and achieve foggy images quickly and effectively addressed. Keywords: Dark Channel prior; Dehazing ; C ;Atmospheric scattering model; 目 次 1 绪论 . 1 1.1 课题背景 .. 1 1.2研究意义 2 1.3 国内外现状 2 1.4 论文主要内容 . 3 2 相关理论及技术基础 .. 5 2.1雾天图像降质原因 . 5 2.2暗原色先验 . 6 2.3滤波技术 6 2.4大气散射退化模型 . 7 2.5基于 opencv下的 C语言编程 . 8 3 基于暗原色的雾天图像增强算法 . 12 3.1暗原色先验去雾 .. 12 (责任编辑：qin) |