摘要
介绍在3×3小矩阵中直接采用中心像素值与周围各像素值分别求差,再分别将各个差值与阈值相比较,最后将各个结果综合起来判断此像素点是否为边缘点的方法。判断过的像素点对当前阈值进行了一定的修改,使之随实际情况上下浮动,更有利于当前像素点的判断。实验证明,该方法检测出的边缘更完整,噪声也能更有效地去除。
This article introduces a method of finding margin directly using central image element and peripheral image elements in 3×3 small matrix and then respectively compares the margin of central and peripheral image elements with threshold value. Before doing this we have modified the threshold value based on the judged image elements. Facts have proved that this kind of method can measure the edge of facial image more integrally and remove the yawp more effectively.
出处
《山东科技大学学报(自然科学版)》
CAS
2004年第2期65-67,共3页
Journal of Shandong University of Science and Technology(Natural Science)
关键词
小矩阵
像素值差
边缘
阈值
small matrix
margin of image element
edge
threshold value