摘要由于雾天情况下图像的对比度低，图像清晰度不高，给人们的生产和生活带来了严重的影响。

目前对于雾天图像的增强主要从以下两个方面来进行。一种是从图像退化的原因出发，建立大气模型，从而得到恢复后的图像。另一种是从传统的增强图像对比度出发，不考虑图像退化的原因，只将图像的对比度增强。59999

本文主要从增强图像对比度的角度出发，分析了增强对比度的常用方法，并且将目前流行的Retinex算法进行了仿真和总结。首先，本文分析了针对灰度直方图来提高对比度的方法，包括对比度拉伸，全局直方图均衡化，局部直方图均衡化，POSHE算法。然后分析了Retinex算法，包括单尺度（SSR）和多尺度（MSR）算法。经过对比和分析，Retinex算法有较好的处理效果。

毕业论文关键词：图像清晰  直方图均衡化  Retinex

毕业设计说明书（论文）外文摘要

Title      CCD demisting Algorithm   

Abstract

As the case of fog low image contrast, image resolution is not high, given people's production and life has brought a serious impact.

At present, for foggy image enhancement mainly from the following two aspects. One is the reason for the departure from the image degradation, establish atmospheric models, so that the image is restored . Another is to enhance image contrast from the traditional way, without regard to the cause of image degradation, but enhance the image contrast, to facilitate observation of the human eye.

This article from the perspective of enhancing image contrast, analysis of the commonly used methods of contrast enhancement.And the currently popular Retinex algorithm for the simulation and summary. Firstly, the paper analyzes the histogram for ways to improve the contrast, including contrast stretching, global histogram equalization, local histogram equalization, POSHE algorithm, and the processed image are compared with before treatment. Retinex algorithm then analyzes，including single-scale (SSR) and multi-scale (MSR) algorithm. After comparison and analysis, Retinex algorithm has a better treatment effect.

Keywords   Image clarity   Histogram equalization  Retinex

目 次

1  绪论 1

1.1 课题研究背景和研究目的和意义 1

1.2 国内外发展现状 2

1.2.1 基于图像增强的雾天降质图像清晰化技术研究现状 2

1.2.2 基于模型复原的雾天降质图像清晰化研究现状 3

1.3本论文的章节安排 4

2  传统的图像增强算法 5

2.1概述 5

2.2 对比度拉伸 5

2.3 线性变换 6

2.4 分段线性变化 6

2.5 非线性变化 7

2.6 直方图修正 9

2.6.1 灰度直方图 9

2.6.2 直方图均衡化 9

2.6.3 直方图规定化 11

2.6.4 局部直方图均衡化 11

2.6.5 POSHE算法 13

2.6.6 插值自适应直方图均衡化 17

2.7 同态滤波 19

3  Retinex算法