灰色关联亚像素检测理论在油井管套损检测中的应用研究

发布时间：2018-08-28 08:31

【摘要】：随着油气田开发难度的不断增加,油井管套损也愈加严重,对套损的检测和预防日益成为油气田研究领域的热点之一。井下电视系统是进行井下工况监测及套损预防的重要手段之一,通过井下摄像机直接对套管内壁进行摄像,对实时截取到的套损图像进行边缘检测及相应的定量分析可以便捷的判断出套损程度。灰色关联分析作为灰色理论重要的组成部分,已广泛应用于图像工程中。本文尝试将灰色关联分析亚像素检测理论首次引入到套损检测分析中,实现了对套管损坏程度的定量解释。论文研究了灰色关联分析理论,主要介绍了灰色关联分析的定义、数据序列预处理和灰色关联度的计算。针对从井下电视系统截取到的含噪声的套损图像,引入灰色B型关联度算法对油井管套损图像进行去噪声处理。对除噪后的套损图像,首先,提出了采用灰色关联分析算法对油井管套损图像进行边缘检测,实现了图像边缘的像素级定位,解决了传统边缘检测算子对不同特性边缘所表现出来的差异性问题。其次,利用基于Zernike矩的亚像素检测理论对像素级边缘点进一步精确定位,实现了对这些有效边缘点的亚像素级边缘提取,提高了后续套损分析中定量解释的精确度。本文将灰色关联分析和亚像素检测理论相结合,将其应用于油井管套损检测,仿真实验表明该方法具有较好的检测效果,在提高检测精度的同时也减少了运算时间。最后,采用面向对象语言开发了油井管套损检测系统,实现了油井管套损程度的定量解释和相关套损数据的实时显示,为套损检测和预防提供了一定指导作用。
[Abstract]:With the increasing difficulty of oil and gas field development, casing damage of oil well pipe is becoming more and more serious, and the detection and prevention of casing damage has become one of the hotspots in the field of oil and gas field research. Downhole television system is one of the important means to monitor downhole working conditions and prevent casing damage. The edge detection and quantitative analysis of the real time captured casing damage image can easily judge the casing damage degree. As an important part of grey theory, grey correlation analysis has been widely used in image engineering. This paper attempts to introduce the theory of sub-pixel detection of grey correlation analysis into casing damage detection for the first time, and realize the quantitative interpretation of casing damage degree. This paper studies the theory of grey correlation analysis, mainly introduces the definition of grey correlation analysis, the preprocessing of data sequence and the calculation of grey correlation degree. Aiming at the noise-contained casing damage image intercepted from downhole television system, grey B-type correlation algorithm is introduced to remove noise from casing damage image of oil well pipe. First of all, the gray correlation analysis algorithm is used to detect the edge of casing damage image of oil well pipe, and the pixel level location of image edge is realized. It solves the problem of the difference of the traditional edge detection operator to different characteristic edges. Secondly, the sub-pixel detection theory based on Zernike moment is used to locate the edge points of pixel level, and the sub-pixel edge detection of these effective edge points is realized, which improves the accuracy of quantitative interpretation in the subsequent casing damage analysis. In this paper, grey correlation analysis is combined with sub-pixel detection theory, which is applied to oil well casing damage detection. The simulation results show that the method has better detection effect, and improves the detection accuracy and reduces the calculation time. Finally, an oil well casing damage detection system is developed by using object-oriented language. The quantitative interpretation of casing damage degree and the real-time display of relevant casing damage data are realized, which provides some guidance for casing damage detection and prevention.
【学位授予单位】：西安石油大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TE931.2;TP391.41

【参考文献】