In this paper, we have introduced an ultrasonic scan system and the pertinent calculations of the system parameters, for the purpose of nondestructive testing of oil pipes. We have investigated also the technical characteristics of subsystems to be used and implemented in an industrial smart ultrasonic flaw detector system. Different sections of an ultrasonic system, including ultrasonic, environmental and positional parameters sensors have been presented. We have introduced and described in detail the internal processing, which involves a set of operations inside the pipe. We have also presented thorough feature extraction and complementary processing sections of the ultrasonic scan system, such as noise reduction, enhancement, and smoothing that is applied to images. Using neural network techniques, we have conducted the software simulations to classify the processed images of pipe walls. Decision-based; i.e. DBNN and FDNN, are also employed to enhance the boundaries between the corroded and undamaged regions in the pipeline image.
Application of neuro-fuzzy techniques in oil pipeline ultrasonic nondestructive testing
Anwendung von Neuro-Fuzzy-Verfahren auf die zerstörungsfreie Ultraschallprüfung von Ölpipelines
NDT&E International ; 38 , 8 ; 643-653
2005
11 Seiten, 35 Bilder, 2 Tabellen, 22 Quellen
Article (Journal)
English
Neuro-fuzzy approaches for pipeline condition assessment
Tema Archive | 2007
|Ultrasonic Holography in Nondestructive Testing
SAE Technical Papers | 1970
|Neuro-Fuzzy Techniques for Traffic Control
British Library Conference Proceedings | 1997
|Neuro-fuzzy techniques for traffic control
Tema Archive | 1998
|Application of infrared techniques to nondestructive testing of tires
Tema Archive | 1973
|