架空线路弧垂和绝缘子倾角的图像测量方法

发布时间：2018-03-10 11:29

本文选题：架空线路弧垂　切入点：绝缘子倾角　出处：《华北电力大学》2017年硕士论文　论文类型：学位论文

【摘要】：导线弧垂和绝缘子倾角是架空线路的两个重要机械结构参数,为了保证电网安全可靠运行,需要定期对其进行检测。本文将计算机视觉技术和图像处理技术应用到输电线路巡检中,提出了架空线路弧垂和绝缘子倾角的图像测量方法。主要研究内容及成果如下:第一,基于链码跟踪理论,实现了单根导线的提取,采用形态学方法对边缘检测图像进行处理,有效消除了导线双边缘和断裂现象,并通过控制链码跟踪的搜索方向,实现了导线交叉情况下单根导线的自动提取。第二,结合三维DLT和二维DLT理论实现了弧垂测量,采用三维DLT方法对投影矩阵进行求解,然后采用二维DLT方法计算导线序列点的空间二维坐标。试验结果表明,测量结果与实际情况相符,具有较高的测量精度,达到了工程实际应用的要求。第三,对单目标的金字塔匹配策略进行改进,实现了多目标的自动识别提取,并根据绝缘子串的形状特征对绝缘子进行识别、提取。本文将形态学处理后的二值图像和原二值图像进行“与”运算,解决了绝缘子失真的问题。此外,提出了一种针对绝缘子端点的图像特征匹配方法:选取绝缘子串首尾伞裙与芯棒结合部分中心点作为绝缘子端点,实现了两幅图像中绝缘子端点的精确匹配。第四,结合单目视觉理论,提出了一种绝缘子倾角的图像测量方法,首先使用标定好的相机在不同位置采集两幅目标图像,然后利用单目视觉理论知识计算绝缘子上对应两点的空间三维坐标,进而计算绝缘子空间倾斜角。试验结果表明该方法操作简单,易于实现,测量精度高。
[Abstract]:Wire sag and insulator inclination are two important mechanical structural parameters of overhead lines, in order to ensure the safe and reliable operation of the power network, In this paper, computer vision technology and image processing technology are applied to the inspection of transmission lines, and an image measurement method of overhead line sag and insulator inclination is proposed. The main research contents and results are as follows: first, Based on chain code tracking theory, the extraction of single wire is realized, and the edge detection image is processed by morphological method, which effectively eliminates the phenomenon of double edges and breakage of wire, and controls the search direction of chain code tracking. The automatic extraction of the root wire is realized. Secondly, the sag measurement is realized by combining the three-dimensional DLT and two-dimensional DLT theory, and the projection matrix is solved by using the three-dimensional DLT method. Then the two-dimensional DLT method is used to calculate the two-dimensional coordinates of the traverse sequence points. The experimental results show that the measurement results are consistent with the actual situation and have a high measurement accuracy, which meet the requirements of practical engineering application. The pyramid matching strategy of single target is improved to realize the automatic recognition and extraction of multi-target, and the insulator is recognized according to the shape feature of insulator string. In this paper, the binary image and the original binary image are processed by "sum" operation, which solves the problem of insulator distortion. In this paper, a method of image feature matching for insulator endpoints is proposed: the center point of the combination of the insulator string umbrella skirt and the mandrel is selected as the insulator endpoint, and the accurate matching of the insulator endpoint in the two images is realized. 4th, Based on the theory of monocular vision, an image measurement method of insulator inclination angle is proposed. Firstly, two target images are collected at different positions by a calibrated camera. Then the three-dimensional coordinates corresponding to two points on the insulator are calculated by using the monocular vision theory knowledge, and then the spatial tilt angle of the insulator is calculated. The experimental results show that the method is simple, easy to realize and has high measurement accuracy.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TM755

【参考文献】