南京理工大学学报（自然科学版）2017，Vol.41Issue(3)：307-312,6.DOI:10.14177/j.cnki.32-1397n.2017.41.03.006
基于置信区间的自适应加权均值滤波算法
Adaptive weighted mean filtering algorithm based on confidence interval
摘要
Abstract
An adaptive weighted mean filtering algorithm based on a confidence interval is proposed to improve the results of filtered images.The weighted means of the pixels in a filtering window and within the confidence interval are calculated according to the characteristics of Gaussian noise and its effect on an original image.A weighted coefficient is obtained by the linear weighted sum of the gray measure factor and distance measure factor,and the gray correlation and distance correlation are taken into consideration.Finally,the gray of the weighted mean filtered image is equalized.The experimental results show that this algorithm is better than the standard mean filtering(SMF) algorithm and adaptive mean filtering(AMF) algorithm,the filtered image is clearer,the original image is recovered well,and the edges and details are kept;the normalized mean square error(NMSE) of this algorithm is lower than that of the SMF and AMF.关键词
置信区间/均值滤波算法/自适应滤波算法/高斯噪声/灰度相关性/距离相关性Key words
confidence interval/mean filtering algorithm/adaptive filtering algorithm/Gaussian noise/gray correlation/distance correlation分类
信息技术与安全科学引用本文复制引用
陈家益,黄楠,熊刚强,曹会英,徐秋燕..基于置信区间的自适应加权均值滤波算法[J].南京理工大学学报（自然科学版）,2017,41(3):307-312,6.基金项目
国家自然科学基金(61170320) （61170320）
广东省自然科学基金(2015A030310178) （2015A030310178）
