{$cfg_webname}
主页 > 电子信息 > 电子 >

数字图像增强及去噪技术的研究与应用

来源:wenku163.com  资料编号:WK16313851 资料等级:★★★★★ %E8%B5%84%E6%96%99%E7%BC%96%E5%8F%B7%EF%BC%9AWK16313851
资料介绍

数字图像增强及去噪技术的研究与应用(任务书,开题报告,论文15000字,代码)
摘  要
对于低光照和低对比度图像,我们往往希望知道图像的细节信息。本文通过对基于直方图增强理论和技术的研究与分析,选择基于模糊逻辑的增强算法来处理低光照彩色图像。该算法分为模糊、隶属值修改、去模糊三部分,基于两个关键性参数,均值M和对比强度K。经实验验证,该算法在高亮度区域存在过增强的缺陷,本文通过调整去模糊函数,将对比强度K与图像均值和方差建立联系对算法进行改进,并结合白平衡与量化技术增强图像,实验证明改进的算法效果显著。本文还将两种增强算法相结合,对图像进行二次增强,实验结果显示该做法可改善增强效果。
图像在增强后,其包含的噪声信号也随之增强。本文采用基于DCT变换分块去噪算法进行去噪。将图像分块,对每块图像作DCT变换,阈值处理后再反变换回来。最后每个像素点的值由重叠块中像素点值加权平均得到。实验显示该算法去噪效果显著。
本文最后结合以上增强和去噪算法,完成增强及去噪应用系统的设计,并对低照度环境下实际拍摄的图像画面进行处理。处理结果显示该系统可以较好地改善低光照图像的显示效果,在实际需求中,可用于对夜间视频监控画面进行处理。
关键词:图像增强;图像去噪;模糊逻辑;白平衡;DCT变换

Abstract
For low bright and low contrast images,we always want to know the details of them.By discussing and analyzing the image enhancement algorithms based on histogram,A image enhancement method based on fuzzy-logic and histogram is chosen to process low bright color images.The algorithm consists of three stages: image fuzzification, modification of membership values and image defuzzification, based on two vital parameters M and K,where M is the average intensity value of the image,and K is the contrast intensification parameter.This method has drawbacks in process the high bright region of images.To improve the algorithm,this paper adjusts the function of modification of membership values,makes K related to the mean and the variance of images and combines the technology of color balance and quantification , which is shown significant performance in experiments.This paper also combines the two enhancement algorithm to make images enhanced twice.And results of experiments show it can improve the enhance effect.
After the enhancement of an image,its noisy signals would be enhanced as well.In this paper, the denoising method based on DCT transformation is used.First dividing the image into patches, for each patch,make DCT transformation and threshold processing,then make DCT inverse transformation.Finally, each pixel value is the aggregation and average of pixels’ value in overlapping blocks.Experiments shows this algorithm has a remarkable denoising effect.
This paper combines the above discussed techniques to design an images enhancement and denoising system.Then to process the low bright photos taken at night with the system.The process results show that this system can improve the visual performance of low bright photos,and can be applied to monitoring pictures at night.
Key Words:image enhancement;image denoising;fuzzy-logic;color balance;DCT transformation
 
目  录
摘要..................................................................................................................................................I
Abstract...........................................................................................................................................II
第1章 绪论    1
1.1 背景意义    1
1.2 图像增强及去噪研究现状    1
1.2.1 图像增强算法概述    1
1.2.2 图像去噪算法概述    2
1.2.3 图像质量评价指标    3
1.3 本文的主要工作    3
1.4 本文的结构安排    4
第2章 基于直方图变换的图像增强算法    5
2.1 限制对比度自适应直方图均衡化增强算法    5
2.1.1 CLAHE增强原理    5
2.1.2 算法实现步骤    6
2.1.3 实验结果及分析    6
2.2 基于模糊逻辑和直方图增强算法    8
2.2.1 模糊逻辑原理    8
2.2.2 算法描述及实现步骤    8
2.2.3 实验结果及存在的问题    9
2.3 本章小结    11
第3章 基于模糊逻辑的自适应增强算法    12
3.1 算法描述    12
3.1.1 参数分析    12
3.1.2 去模糊函数修正    12
3.1.3 量化    13
3.1.4 实验结果与分析    14
3.2 色彩平衡    15
3.2.1 色彩平衡原理及实现    16
3.2.2 实验结果与分析    16
3.3 算法流程图    17
3.4 实验结果与分析    17
3.5 改进算法结合CLAHE的二次增强    19
3.6 本章小结    20
第4章 彩色图像增强及去噪技术的应用    21
4.1 图像去噪    21
4.1.1 块匹配去噪    21
4.1.2 基于DCT域分块去噪    23
4.2 彩色图像增强及去噪的应用    26
4.2.1 应用系统框架    26
4.2.2 实验结果及分析    26
4.3 本章小结    28
第5章 总结与展望    29
5.1 全文总结    29
5.2 下一步工作展望    29
参考文献    30
致  谢    32
 

推荐资料