详细信息

图像修复中物体移除的实验设计    

The Design of Experiment on Objection Moving in Image Inpainting

文献类型:期刊文献

中文题名:图像修复中物体移除的实验设计

英文题名:The Design of Experiment on Objection Moving in Image Inpainting

作者:于文静[1]

机构:[1]华东理工大学电子与信息实验教学中心,上海200237

年份:2013

卷号:32

期号:12

起止页码:126

中文期刊名:实验室研究与探索

外文期刊名:Research and Exploration In Laboratory

收录:CSTPCD;;北大核心:【北大核心2011】;

语种:中文

中文关键词:图像修复;高斯滤波;边缘检测;小波变换

外文关键词:image inpaintin; gaussian filtering; edge detection; wavelet transformation

摘要:设计了图像修复中物体移除的实验,通过实验的操作激发学生对图像修复进行研究的兴趣。通过此实验,学生可以学会高斯滤波的使用,各种边缘提取的方法及区别,利用小波变换和边缘检测获得图像的主要结构,在物体移除后从结构和纹理两个方面实现图像的修复。首先对图像修复中物体移除的基本原理及实现过程进行解释,然后利用这一方法实现物体移除后破损图像的修复。提供可实现上述方法的Matlab程序代码,学生可利用相关程序对各种物体移除图像进行处理,得到了视觉上无明显人工修复痕迹的图像。在数字图像实验中首次引入了结构和纹理分别修复的图像修复方法。
The experiment on objection moving in image inpainting is designed to deepen students' understanding about the concept and the implementation of the image inpainting technology. The experiment stimulates students' interest in research on the image inpainting technology, trains students to handle the use of Gaussian filtering, know the difference in all kinds of edge detection, learn to achieve the salient structure by use of wavelet transform and edge detection, and win image inpainting after objection moving in structure and texture. Firstly, the experiment explains the basic principle and the implementation of this method. Moreover, objection moving damaged image are inpainted by this method. This essay also provides the matlab code which could achieve objection moving image inpainting so that students could use relative programs to restore different types of damaged images and get images with no obvious artificial traces visually. Moreover, students could get different inpainting results through changing the different damaged images so as to deepen their understanding the method by analyzing the results. The based on structure and texture image restoration method is implemented in digital image experiments for the first time.

参考文献:

正在载入数据...

版权所有©华东理工大学 重庆维普资讯有限公司 渝B2-20050021-7 
渝公网安备 50019002500408号 违法和不良信息举报中心