电子学报 ›› 2016, Vol. 44 ›› Issue (4): 878-885.DOI: 10.3969/j.issn.0372-2112.2016.04.018

• 学术论文 • 上一篇    下一篇

基于不确定性信息融合的高密度椒盐噪声降噪方法

齐现英1,2, 刘伯强1, 徐建伟3   

  1. 1. 山东大学控制科学与工程学院, 山东济南 250061;
    2. 泰山医学院放射学院, 山东泰安 271000;
    3. 泰安市肿瘤防治院影像科, 山东泰安 271000
  • 收稿日期:2015-04-17 修回日期:2015-06-29 出版日期:2016-04-25
    • 通讯作者:
    • 刘伯强
    • 作者简介:
    • 齐现英 女,1974年7月出生,山东泰安人.2003年于山东科技大学信息科学与工程学院获硕士学位.现为山东大学博士研究生,研究方向为医学图像处理及医学仪器. E-mail:xyqi@tsmc.edu.cn
    • 基金资助:
    • 国家自然科学基金 (No.61203330)

A Novel Algorithm for Removing High-Density Salt-and-Pepper Noise Based on Fusion of Indeterminacy Information

QI Xian-ying1,2, LIU Bo-qiang1, XU Jian-wei3   

  1. 1. School of Control Science and Engineering, Shandong University, Jinan, Shandong 250061, China;
    2. Department of Radiology, Taishan Medical University, Taian, Shandong 271000, China;
    3. Taian Cancer Prevention and Treatment Hospital, Taian, Shandong 271000, China
  • Received:2015-04-17 Revised:2015-06-29 Online:2016-04-25 Published:2016-04-25
    • Supported by:
    • National Natural Science Foundation of China (No.61203330)

摘要:

为解决高密度椒盐噪声滤除与细节保护之间的矛盾,提出一种基于不确定性信息融合的中智灰滤波算法.该算法包括两个阶段:噪声检测和噪声恢复.在检测阶段,为提高噪声检测准确率,首先利用Max-Min算法进行初测,然后利用极值压缩灰色关联度与顺序不确定性的融合信息进行二次判断.在噪声恢复阶段,为充分利用像素本身的不确定性及邻域像素的灰色关联性,将中智不确定性和极值压缩灰色关联度的乘积作为相似性度量特征,设计了中智灰自适应权重函数.实验表明,针对不同图像,二次噪声检测方案的噪声剔除率可达0.1%~8.8%;该中智灰滤波算法在抑制椒盐噪声的同时能较好地保护图像边缘信息,特别是在高噪声(70%~90%)情况下,算法的综合性能优于现有相关算法.

关键词: 高密度椒盐噪声, 二次噪声检测, 中智灰自适应权重, 极值压缩灰色关联度, 顺序不确定性, 中智理论

Abstract:

To solve the contradiction of image denoising and detail-preserving under high-density salt-and-pepper noise,this paper proposes a Neutrosophy-Gray filter by using the fusion of indeterminacy information.It has a two-stage scheme:noise detecting and noise removing.In detecting stage,to improve the accuracy of noise detection,Max-Min algorithm is employed firstly,then noise candidates are judged again by the dual criteria of Extreme-Compression-Grey-Correlation-Degree (ECGCD) and Ordered-Indeterminacy (OI).In filtering stage,the algorithm applies the multiplicative fusion of ECGCD and indeterminacy to measure the similarity of pixels,and a Neutrosophy-Gray adaptive weighted function is presented.Experiments show,for different images,the rates of noise eliminating change between 0.1% and 8.8%,and performances of denoising and detail-preserving of the proposed algorithm are superior to current filters even at high level noise (70%~90%).

Key words: salt-and-pepper noise with high-density, double noise detection, neutrosophy-gray adaptive weight, extreme-compression-grey-correlation-degree, ordered-indeterminacy, neutrosophy

中图分类号: