基于阈值分割的叶片缺陷检测技术  

Detection of Vane Defects Based on Otsu Thresholds

作　　者：吴清[1] 赵歆波[1] 王庆胜[1] 

机构地区：[1]西北工业大学现代设计与集成制造技术教育部重点实验室,西安710072

出　　处：《机床与液压》2006年第9期178-180,共3页Machine Tool & Hydraulics

基　　金：航空科学基金资助(项目编号04I53069)

摘　　要：针对CT缺陷图片的灰度特点提出了一种基于二维O tsu阈值分割缺陷检测方法。详细介绍了该方法的理论依据和实现过程,并给出了具体的解决方案。该方法检测铸件缺陷准确率较高、可靠性好、速度快、可实现自动检测,为CT检测提供了一种行之有效的方法。A method of defects recognition using the techniques of planar Otsu thresholds aiming at the characteristic of the CT defect image was introduced. The theory and the process of realization were described. The resolve was presented. This method can get the high veracity, good reliability, rapid speed and effective result . It is an effective method for CT.

关 键 词：ICT 缺陷检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]