灰度等级处理中的OSTU动态阈值法研究  

Study of OSTU Dynamic Threshold Based on Gray-scale Processing

作　　者：张辉[1] 张道勇 何最红[3] 

机构地区：[1]广东教育学院物理系电子教研室 [2]广州数控设备有限公司 [3]广州大学实验中心

出　　处：《传感器世界》2008年第7期24-27,共4页Sensor World

摘　　要：通过对PCB图像光学特性的分析,提出一种自适应中值加权均值混合滤波器,针对CCD采集的图像,进行预处理;并在此基础上,提出了一种基于灰度等级处理的OSTU动态图像二值化算法,从提高运算速度和分割精度两方面对最大类间方差法进行了改进,增强了图像分割的抗噪声能力,有效的改善了图像的分割效果。An adaptive median-weighted mean hybrid filter is advanced based on the analysis of PCB image optical characteristics and pretreatment is carried out aimed at CCD collected image. On this basis, a method of OSTU dynamic image binarization algorithm based on gray-scale processing is also put forward. In order to accelerate the computational speed and improve the precision segmentation, maximum between-class variance method has been improved, which has enhanced the image segmentation＇s anti-noise capability and effectively improved the image＇s segmentation results.

关 键 词：灰度等级 动态阈值 加权均值 自适应中值 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]