摘要：【中文】本发明公开了一种复合绝缘子内部粘接缺陷的检测方法和系统，通过激光阵列设备、热激励光源设备、成像设备获取被检测复合绝缘子的可见光图像、第一散斑图像和第二散斑图像，通过数据处理设备对第一散斑图像和第二散斑图像进行运算处理，得到干涉条纹图和相位图，再进行高斯滤波处理，绘制边缘图像与可见光图像融合，得到融合图像，识别复合绝缘子内部缺陷。本发明实现对复合绝缘子内部粘接缺陷快速、精确、无接触、直观的检测，对复合绝缘子内部微小的粘接缺陷也具有良好的检测效果，实现高效精准地检测复合绝缘子内部粘接缺陷的目的，本发明操作简单，成像直观，受检测人员的操作主观因素的影响较小，能有效保障检测的客观性和科学性。 【EN】The invention discloses a method and a system for detecting bonding defects inside a composite insulator. The invention realizes the rapid, accurate, non-contact and visual detection of the bonding defect inside the composite insulator, has good detection effect on the tiny bonding defect inside the composite insulator, and realizes the purpose of efficiently and accurately detecting the bonding defect inside the composite insulator.

摘要：【中文】本发明提供一种GIS管道缺陷在线检测装置，包括检测装置、控制系统和数据分析处理系统，检测装置主要包括闪光灯热激励单元和红外热像采集单元；控制系统与检测装置通信连接，控制系统用于调节闪光灯热激励单元的激励产生时间及频率，控制系统还用于调节红外热像采集单元的采集频率和采集时间；数据分析处理系统与检测装置以及控制系统均通信连接，数据分析处理系统采用频域热特征成像方法对数据进行分析处理。本发明能够快速、准确、定量地检测出存在缺陷的GIS管道，并且能够多角度地收集红外图像序列，实现对GIS管道缺陷的位置、深度、大小的识别。本发明还提供一种应用在该检测装置中的GIS管道缺陷在线检测方法。 【EN】The invention provides a GIS pipeline defect online detection device which comprises a detection device, a control system and a data analysis processing system, wherein the detection device mainly comprises a flash lamp thermal excitation unit and an infrared thermal image acquisition unit; the control system is in communication connection with the detection device, is used for adjusting the excitation generation time and frequency of the flash lamp thermal excitation unit, and is also used for adjusting the acquisition frequency and acquisition time of the infrared thermal image acquisition unit; the data analysis processing system is in communication connection with the detection device and the control system, and the data analysis processing system analyzes and processes data by adopting a frequency domain thermal characteristic imaging method. The method can quickly, accurately and quantitatively detect the GIS pipeline with defects, and can collect infrared image sequences at multiple angles to realize the identification of the positions, depths and sizes of the defects of the GIS pipeline. The invention also provides a GIS pipeline defect online detection method applied to the detection device.

摘要：【中文】本发明的确定疑似家族性缺陷的方法，属于电力系统的技术领域，解决现有技术中的方法不能智能化利用缺陷记录数据为电网系统服务的技术问题。该方法包括获取电网系统中的电网设备数据，所述电网设备数据包括：设备基础台账数据、设备部件数据和设备分类数据；根据所述分类数据建立所述基础台账数据与所述设备部件数据的对应关系；获取电网系统中已知缺陷电网设备的第一数据；确定所述对应关系中的未知电网设备中与所述第一数据相同的数据，并提取所述相同的数据。本发明用以电力系统的功能，满足人们对电力系统工作稳定和高效要求。 【EN】The invention discloses a method for determining suspected familial defects, belongs to the technical field of power systems, and solves the technical problem that the method in the prior art cannot intelligently utilize defect record data to serve a power grid system. The method comprises the steps of obtaining power grid equipment data in a power grid system, wherein the power grid equipment data comprise: device base ledger data, device component data, and device classification data; establishing a corresponding relation between the basic standing book data and the equipment component data according to the classification data; acquiring first data of known defective power grid equipment in a power grid system; and determining the data which are the same as the first data in the unknown power grid equipment in the corresponding relation, and extracting the same data. The invention is used for the functions of the power system and meets the requirements of people on the stable and high-efficiency work of the power system.

摘要：【中文】本申请实施例提供了一种电力设备铭牌图像的文本检测方法及装置，其中方法包括：获取输入图像；利用深度学习算法及边缘检测算法，对所述输入图像中的铭牌进行定位提取；通过透视变换对定位提取图像进行形变矫正，得到矫正后的铭牌图像；利用深度学习文本检测算法，对矫正后的铭牌图像自动进行文本检测，得到铭牌文本检测结果。本申请的方案通过深度学习算法实现对电力设备铭牌图像中的文本的智能检测，解决了电力设备信息自动采集中铭牌文本难以检测的难题，提高了电力设备信息采集的效率。 【EN】The embodiment of the application provides a text detection method and a text detection device for a nameplate image of electric power equipment, wherein the method comprises the following steps: acquiring an input image; positioning and extracting a nameplate in the input image by utilizing a deep learning algorithm and an edge detection algorithm; carrying out deformation correction on the positioning extracted image through perspective transformation to obtain a corrected nameplate image; and automatically carrying out text detection on the corrected nameplate image by utilizing a deep learning text detection algorithm to obtain a nameplate text detection result. According to the scheme, intelligent detection of the text in the nameplate image of the power equipment is achieved through the deep learning algorithm, the problem that the nameplate text is difficult to detect in automatic collection of the information of the power equipment is solved, and the information collection efficiency of the power equipment is improved.