摘要
针对现行的均值滤波算法存在的局限性,基于灰度修剪和均衡化的加权均值滤波算法对其进行改进.算法根据高斯噪声的特点及其对原图像的影响,对处于灰度概率峰值附近所对应的灰度进行修剪,再进行加权均值滤波.加权系数同时考虑灰度相关性与距离相关性,是灰度测度因子和距离测度因子的乘积.算法最后对加权均值滤波后图像进行分段的灰度均衡化.滤波实验的结果表明,相对于现行的均值滤波算法,本算法有着更好的滤波性能,在滤除噪声的同时,很好地保持图像的边缘和细节部分.
Against the limitation of existing mean filtering algorithms,an improved weighted mean filtering algorithm is proposed by gray trimmed and equalization. According to Gaussian noise characteristics and its effect on original image,the corresponding gray is firstly trimmed to gray probability peak,and then the noise image is filtered by weighted mean. The weighted coefficient is the product of gray measure factor and distance measure factor,which takes gray correlation and distance correlation into consideration. Finally,the algorithm piecewise equalizes the image gray of weighted mean filtered. Experimental results demonstrate that the proposed algorithm has a significant better filtering performance in comparison with the existing mean filtering algorithms,which maintains image edges and details well in filtering noise.
引文
[1]TANG W,TANG J.An improved mean filter algorithm for impulse noise elimination[J].Advanced Materials Research,2014,936:2281-2285.
[2]郑秀清,何坤,张健.基于Prewitt算子的TV图像去噪算法[J].四川师范大学学报(自然科学版),2013,36(3):456-462.
[3]武英,吴海勇.一种自适应图像去噪混合滤波方法[J].计算机工程与应用,2010,46(7):168-170.
[4]罗丹.基于图像增强的改进自适应加权均值滤波算法[J].电视技术,2013,37(23):15-22.
[5]杨吉宏,胡顺波,张民,等.灰度图像加权有向平滑滤波算法[J].计算机工程与设计,2010,31(1):153-156.
[6]GONZALEZ R C,WOODS R E.Digital Image Processing[M].北京:电子工业出版社,2010.
[7]SAKTHIVEL N,PRABHU L.Mean-median filtering for impulse noise removal[J].Inter J Basic and Appl Sci,2014,2(4):47-57.
[8]FANG J,XU D C,FU X.An improved hybrid median-mean filter algorithm[J].Appl Mechanics Materials,2015(701/702):288-292.
[9]常瑞娜,穆晓敏,杨守义,等.基于中值的自适应均值滤波算法[J].计算机工程与设计,2008,29(16):4257-4259.
[10]KANG J Y,MIN L Q,LUAN Q X,et al.Novel modified fuzzy C-means algorithm with application[J].Digital Signal Processing,2009,19(2):309-319.
[11]ZHANG P X,LI F.A new adaptive weighted mean filter for removing salt-and-pepper noise[J].IEEE Signal Processing Letters,2014,21(10):1280-1283.
[12]常瑞娜,穆晓敏,杨守义,等.基于局部空间像素特征的自适应加权滤波算法[J].计算机工程与应用,2008,44(6):45-67.
[13]曹萌,张有会,王志巍,等.基于城区距离的自适应加权均值滤波算法[J].计算机应用,2013,33(11):3197-3200.
[14]朱士虎,游春霞.一种改进的均值滤波算法[J].计算机应用与软件,2013,30(12):97-116.
[15]MIAO Z W,JIANG X D.Weighted iterative truncated mean filter[J].IEEE Transactions on Signal Processing,2013,61(16):4149-4160.
[16]刘书,王慈.基于自适应非局部均值滤波的图像去块算法[J].上海交通大学学报,2013,47(12):1930-1933.
[17]李佐勇,汤可宗,胡锦美,等.椒盐图像的方向加权均值滤波算法[J].中国图象图形学报,2013,18(11):1407-1415.
[18]曹洁,李伟.一种高精度自适应粒子滤波算法[J].兰州理工大学学报,2011,37(3):83-88.
[19]彭良玉,杨辉,黄满池.一种新的基于脉冲噪声点检测的自适应中值滤波算法[J].兰州理工大学学报,2009,35(1):77-80.
[20]李玲,胡学刚,蒋伟.一种基于LIP的全变分图像去噪新模型[J].四川师范大学学报(自然科学版),2011,34(1):134-138.
[21]盛骤,谢式千,潘承毅.概率论与数理统计[M].北京:高等教育出版社,2013:46-50.