基于阶乘权值模板的卷积滤波算法  

Convolution filtering algorithm based on factorial weight template

作　　者：董林鹭 杨平先 陈明举 DONG Lin-lu;YANG-Ping-xian;CHENG Ming-ju(Colleag of Information Engineering,Sichuan University of Science&Engineering,Zigong 64300,China)

机构地区：[1]四川理工学院自动化与信息工程学院

出　　处：《电子世界》2018年第17期42-43,共2页Electronics World

摘　　要：针对传统的空域平滑卷积模板的权值系数未有效利用图像相邻空间的位置信息,本文构建一种平滑卷积模板的权值系数由图像相邻距离的阶乘确定的空域滤波方法。通过对比实验证明,本文提出的基于阶乘权值模板的卷积滤波算法更好地地保留图像的细节与轮廓信息,峰值信噪比与结构相似度都有一定的提高,图像去噪性能更优。For the traditional spatial domain smooth convolution template weight coefficient does not use the location information of the image adjacent space effectively,this paper constructs a spatial filtering method that the weight coefficient of the smooth convolution template is determined by the factorial of the image adjacent distance.Comparison experiments show that the convolution filtering algorithm based on factorial weight template proposed in this paper can preserve the details and contour information of the image better,the peak signal-to-noise ratio and structural similarity are improved,and the image denoising performance is better

关 键 词：阶乘 空间域滤波 高斯噪声 图像去噪 

分 类 号：TN911.7[电子电信—通信与信息系统]