| 基于加权核范数最小化的彩色图像去马赛克方法 | |
| Alternative Title | Color Image Demosaicing Based on Weighted Nuclear Norm Minimization |
| 张旭 | |
| Subtype | 硕士 |
| Thesis Advisor | 黄玉梅 |
| 2020-05-17 | |
| Degree Grantor | 兰州大学 |
| Place of Conferral | 兰州 |
| Degree Name | 理学硕士 |
| Degree Discipline | 计算数学 |
| Keyword | 彩色图像去马赛克 加权核范数最小化方法 交替迭代算法 |
| Abstract | ~彩色图像中, 每个像素点的像素值由红、绿、蓝三个颜色成分的值决定, 每种颜 色成分称为一个颜色通道. 然而彩色图像记录设备在记录图像时采用的是单通彩色 滤波阵列, 单通彩色滤波阵列在每个像素点只记录一个颜色成分的值, 例如红色的 值、或者绿色的值、或者蓝色的值. 所以彩色图像记录设备记录到的图像为马赛克 图像, 需要利用图像去马赛克算法对马赛克图像进行处理获得彩色图像. 图像去马 赛克就是利用单通彩色滤波阵列记录到的单色信息, 去获得每个像素点的其他两个 颜色信息, 最后获得一幅彩色图像的过程. 由此可见, 图像去马赛克算法是人们利用 彩色图像记录设备获取彩色图像的重要技术, 所以图像去马赛克算法的研究具有重 要的理论及实际意义. 本论文提出了一个基于加权核范数最小化的彩色图像去马赛克方法. 首先, 介 绍了去马赛克问题并概述去马赛克方法, 简单介绍加权核范数最小化方法其次, 利 用彩色图像颜色通道间的相关性信息, 基于加权核范数最小化方法建立彩色图像去 马赛克模型, 并设计交替迭代算法求解该模型最欢, 给出所构造方法的图像去马赛 克的实验结果，并与现有的其他去马赛克方法所获得的结果进行比较, 视觉和数值 结果显示, 本文所提出的模型得到了效果更好的去马赛克图像. |
| Other Abstract | In color images, the color information at each pixel is represented by three colors, red, green and blue. Each color is called a channel. But the cameras utilize color filter array which just captures one color at each pixel, such as red, green or blue. So the captured image is mosaic image. Then digital cameras need to use the demosaicing algorithm to demosaic and then obtain color image. The demosaicing algorithm is a process of getting color images by utilizing the captured color to obtain the other two colors at each pixel. We can find that the demosaicing algorithm is a very important technique to get color images in digital cameras. Therefore, the research of demosaicing algorithm has important theoretical and practical significance. In this paper, the method we focus on is color image demosaicing based on weighted nuclear norm minimization. Firstly, we introduce the problem of color image demosaicing, give the summary of demosaicing methods and also introduce the weighted nuclear norm minimization method. Secondly, we use inter-channel correlation information to propose the color image demosaicing model which based on weighted nuclear norm minimization method. Thirdly, we design alternating iterative algorithm to solve the proposed model. At last, we list results of our method and some other mthods. We compare our results with other demosaicing methods in two aspects, visual aspect and experimental aspect. The results demonstrate that the demosaiced images obtained by our proposed model are better. |
| Pages | 33 |
| URL | 查看原文 |
| Language | 中文 |
| Document Type | 学位论文 |
| Identifier | https://ir.lzu.edu.cn/handle/262010/463212 |
| Collection | 数学与统计学院 |
| Affiliation | 数学与统计学院 |
| First Author Affilication | School of Mathematics and Statistics |
| Recommended Citation GB/T 7714 | 张旭. 基于加权核范数最小化的彩色图像去马赛克方法[D]. 兰州. 兰州大学,2020. |
| Files in This Item: | There are no files associated with this item. | ||||
