朱明,杨利杰,吕金燕,王梦飞.针对多重退化因素的印刷图像复原方法[J].包装工程,2018,39(19):190-196.
ZHU Ming,YANG Li-jie,LYU Jin-yan,WANG Meng-fei.A Printing Image Restoration Method for Multiple Degradation Factors[J].Packaging Engineering,2018,39(19):190-196. |
| 针对多重退化因素的印刷图像复原方法 |
| A Printing Image Restoration Method for Multiple Degradation Factors |
| 投稿时间：2018-05-31 修订日期：2018-10-10 |
| DOI：10.19554/j.cnki.1001-3563.2018.19.033 |
| 中文关键词: 图像复原 椒盐噪声 高斯噪声 引导滤波 |
| 英文关键词: image restoration salt & pepper noise Gaussian noise guided filter |
| 基金项目:国家自然科学基金（61301231）；2017年度河南省青年骨干教师培养计划（2017GGJS154） |
|
| 摘要点击次数: |
| 全文下载次数: |
| 中文摘要: |
| 目的 对于由多种因素所导致的印刷图像退化问题，文中提出一种针对椒盐噪声、高斯噪声和模糊退化等多重退化因素的图像复原方法。方法 首先针对印刷图像椒盐噪声密度不高的特点，设计一种基于灰度范围准则和局部差别准则的椒盐噪声二级检测和滤除方法，并通过评价实验得出合适的阈值参数设置。在去除高斯噪声和图像模糊的过程中，利用边缘保持平滑滤波的原理和特性，将双边滤波器和引导滤波器应用于图像复原中，又在此基础上设计和应用图像细节增强的二次引导滤波器。结果 在椒盐噪声去除方面，新方法对大部分图像都能取得较好的复原效果，尤其对细微边缘不多的图像效果最佳，复原后的PSNR值能达到40以上。二次引导滤波器对高斯噪声和图像模糊的复原效果最好。结论 通过对不同图像复原方法的效果进行评价和分析，验证了文中方法的性能，为今后图像复原技术的应用提供了指导。 |
| 英文摘要: |
| The work aims to propose an image restoration method for printing image degradation caused by multiple factors with respect to multiple degradation factors including salt & pepper noise, Gaussian noise and fuzzy degradation. First, considering that the salt & pepper noise density of printing image was not very high, a two-step method for detecting and removing salt & pepper noise based on gray-scale range criterion and local difference criterion was designed, and the proper threshold parameters were found by means of the evaluation experiment. In the process of removing Gaussian noise and image blurring, considering the principle and properties of the edge-reserving smoothing filter, the bilateral filter and guided filter were applied in the image restoration, and on this basis, the second-time guided filter for detail-enhancement of the image was designed and applied. For the removal of salt & pepper noise, the proposed method can achieve better restoration results for most images, especially for the images with few slight edges, and the PSNR value can reach over 40 after the restoration. The second-time guided filter had the best restoration effect for Gaussian noise and image blurring. Through the evaluation and analysis on different image restoration methods, the properties of the proposed method are verified and it provides guidance for the application of image restoration technology in the future. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
渝公网安备 50010702501716号