基于多幅图像平均法对数字图像降噪的研究  

Study on digital image noise based on many images average method

作　　者：马艳娥[1] 陈思[1] 陈娟[1] 刘妍[1] 姚金杰[1] 

机构地区：[1]中北大学信息探测与处理技术研究所,太原030051

出　　处：《电子测试》2011年第6期1-3,15,共4页Electronic Test

基　　金：国家自然科学基金资助项目(60532080)

摘　　要：针对数字图像在采集和传输过程中,经常会受到噪声的影响,使图像发生降质的这一问题,提出了一种多幅图像平均法。该方法通过对同一景物的多次采集图像相加后取平均值来消除噪声。在实际运算过程中不需要图像的统计特性,可以克服线性滤波器所带来的图像细节模糊,而且对滤波脉冲干扰及图像噪声最有效。实验结果表明这种算法能有效地对图像降噪,并采用客观评价指标可以得出,很大程度提高算法的降噪效果,增强了该算法的实用性。Aiming at the problem of noise impact on digital image in the collection and transmission process,make the image happen drop this problem,a many images average method is proposed,which is applied to multiple acquisition image through adding take average to eliminate noise.In the practical operation process,it does not require image statistical characteristics and it can overcome the image detail fuzzy brought by linear filter,also have the most effective role on filtering pulse disturbance and image noise.Experimental results show that this algorithm can effectively to image noise reduction,and adopt the objective evaluation index can conclude,greatly improving the noise control effect,it can enhance the practicality of the algorithm.

关 键 词：降噪 多幅图像平均法 客观评价指标 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]