自适应控制迭代的随机值脉冲噪声滤波  

Adaptive iterative control for random values impulse noise filtering

作　　者：郭红伟[1,2] 余江[2] 骆洪军[1] 张自红[1] 

机构地区：[1]红河学院工学院,云南蒙自661100 [2]云南大学信息学院,昆明650091

出　　处：《计算机工程与应用》2011年第34期193-195,共3页Computer Engineering and Applications

基　　金：红河学院科研基金(No.XJ1Y0904)

摘　　要：定义了一种更接近实际噪声情况的随机值脉冲噪声模型,针对随机值脉冲噪声的滤除提出一种自适应控制迭代次数的滤波算法。算法包括噪声检测、噪声滤除、误检像素修正和迭代控制四个步骤,对固定值脉冲噪声和随机值脉冲噪声均能有效滤除。与中值滤波算法进行实验比较,在滤除随机值脉冲噪声时,该算法滤波后图像细节信息保护较好,且滤波过程无需设定迭代参数,自适应性强。The random value impulse noise modeling is defined,it is more realistic noise conditions,an adaptive controlling iteration times filter algorithm is proposed for the random value impulse noise remove.Algorithm consists of four steps,noise detection,noise filtering,error detection pixel correction and iterative control.The algorithm can remove fixed-value noise and random-value impulse noise effectively,and can effectively filter out fixed value impulse noise and random value impulse noise.Compared with median value filtering algorithm,this algorithm can protect filtered image information better,and there is no need to set the iteration parameters in the filtering process,with strong adaptability.

关 键 词：脉冲噪声 中值滤波 迭代 自适应性 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]