基于低对比度和小区域模型的快速二值化方法  

Fast Binary Conversion Method for Low-contrast and Small-area Models

作　　者：宗德祥[1] 何永辉[1] 王康健[1] 石桂芬[1] 

机构地区：[1]宝钢研究院设备所,上海201900

出　　处：《宝钢技术》2007年第1期66-68,共3页Baosteel Technology

摘　　要：针对在带钢检测过程中出现的低对比度和小区域特征的缺陷,设计了一种快速高效的图像二值化方法。同OTSU算法和灰度平均值算法比较,该图像二值化方法在具有小区域缺陷特征图像处理方面有一定的优势。该方法对图像进行二值化处理,速度快、精度高,适用于对速度要求较高的图像处理场合。A fast and efficient binary conversion method has been designed to overcome low-contrast and smallarea defects in the strip detecting process. Compared with the Mean Gray Level and OTSU algorithm, the binary conversion method has advantages of processing small-area defect images. As the image binarization can be done quickly and with high accuracy, this method is suitable for places where images are processed at a high speed.

关 键 词：二值化 阈值 机器视觉 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]