Image Recognition Of Magnetic Flux Leakage Inner Detection For Pipeline

Pipelines play an important role in the oil industry, chemical industry and thermoelectricity. It is necessary to inspect and maintain the pipeline regularly that is related to national economic and ecological environment. As a major means of defect detection in pipelines, inner detection has provided reliable scientific protection for safe operation and maintenance of pipelines. Pipeline magnetic flux leakage testing is generally divided into ultrasonic defect detection and defect detection. Magnetic flux leakage defect detection is applied most commonly in China. For a long time, magnetic flux leakage inner detection become the most widely used and most technologically sophisticated means of detection because of its low environment request, no coupler, wide application, low costs and so on.For the problem of defect recognition of magnetic flux leakage pipeline, a method based on threshold analysis is introduced to process MFL defect data. To facilitate image recognition, a series of visual magnetic flux leakage curve is generated. By the use of Delphi programming software, one vertical line can be draw to locate the girth weld automatically in girth weld recognition and one slash in the form of dots can be draw to locate the spiral weld in spiral weld recognition which makes the function of automatic weld recognition basically. Also triangle can be used to mark the position of defect. Also EXCEL for defect has been generated which provides necessary data for defect recognition.A large number of image recognition experiments for magnetic flux leakage data show that the method which is based on threshold analysis of welds and defect feature points can effectively identify the girth welds, spiral welds and defect position in pipelines. EXCEL can accurately output defect location, size and other parameters. The recognition accuracy rate can be above 90%.