复合板灰色自适应瑕疵检测  

Adaptive defect detection of composite plate based on grey system theory

作　　者：杜志顺[1] 吴国平[1] 裘咏霄[1] 黄文丽[1] 陈茂源[1] 

机构地区：[1]中国地质大学机械与电子信息学院,武汉430074

出　　处：《计算机应用》2010年第8期2250-2253,共4页journal of Computer Applications

摘　　要：目标检测是图像处理领域非常关键的一个环节,对实际拍摄的复合板图像进行处理,以灰色系统理论为基础,构建零模灰色关联模型,采用灰度直方图差分自适应法进行动态阈值分割,实现了复合板瑕疵(小目标)的检测。实验表明该方法具有误检率低、稳定性高、自适应能力强、算法简单等特点。Target detection is a key step in image processing.In this paper the adaptive defect detection of composite plate was realized.Firstly The composite plate image was processed by oriented light compensation then a new all-zero grey correlation model based on grey system theory was constructed.And then a finite difference adaptive method using gray histogram was adopted to realize the dynamic threshold segmentation.The theory analysis and real application of this method indicate that it has advantages of low false detection rate high stability strong adaptive ability simple algorithm and so on.

关 键 词：复合板 瑕疵检测 零模灰色关联模型 直方图差分 自适应检测 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论] TP311[自动化与计算机技术—计算机科学与技术]