水下结构物点蚀缺陷的识别与定量分析  

Recognition and quantitative analysis of pitting corrosion defect of underwater structures

作　　者：陆玲[1] 陈国明[1] 林红[1] 

机构地区：[1]中国石油大学机电工程学院,山东东营257061

出　　处：《石油大学学报（自然科学版）》2005年第3期79-82,共4页Journal of the University of Petroleum,China(Edition of Natural Science)

基　　金：国家高科技(863)研究发展专项经费资助项目(2002AA616060)

摘　　要：运行多年的水下结构物会存在不同程度的点蚀缺陷,必须对其进行检测和安全评估。基于图像处理、模式识别以及计算机视觉技术提出了一种水下结构物点蚀缺陷识别与定量分析方法。用采集设备获得点蚀缺陷图像,对其进行特殊的图像处理后,提取其特征并进行识别,通过摄像机系统标定确定缺陷的真实尺寸。分析了水下结构物点蚀缺陷的特征提取和识别的理论依据。用二次标定法对摄像机进行了系统标定,并结合实例给出了标定当量及点蚀特征的实际值。点蚀缺陷识别和定量分析结果为水下结构物应力计算和安全可靠性评估提供了必要的基础数据。The underwater structures may be subject to pitting corrosion defect during their service life. So the regular inspection and safety assessment is necessary. A recognition and quantitative analysis method was presented for pitting corrosion defect of underwater structure based on image processing, pattern recognition and computer vision techonology. The theory basis and realization process of features abstraction and recognition on pitting corrosion defect was discussed in detail. The twice-calibration method was put forward for system calibration of camera. The calibration equivalent and the real dimension of pitting corrosion defect was obtained combining the actual example. The results of recognition and quantitative analysis provide the necessary data for stress calculation and safety assessment of underwater structure.

关 键 词：水下结构物 点蚀缺陷 图像处理 模式识别 计算机视觉技术 应力计算 安全性 海洋石油开发 

分 类 号：TE95[石油与天然气工程—石油机械设备]