Non -destructive inspection of oil pipeline integrity is of primary interes t for an early detec tion of defects that can alter the structure of the pipe. Usual inspection i s performed using a pipeline inspection device equi pped with ultrasonic sensors (up to 512) regularly dispatched all around its circumf erence. The tool is inserted in the pipeline and is driven by the flow of the medium. The travel ling time of the ultrasound beam from the sensor to the internal surface and from the internal surface and the external surface of the pipe is converted i nto distance that reflects the thickness of the pip e versus the sensor angular position, and the discr ete location of the inspection tool along the path. Data are visualized as large images (up to 10 Giga pixels) highly disto rted by intense noise. Processing of these data is made manu ally and is particular ly time consuming. Following previous publications (2011 NDT in Progress and 2012 ICNDT conference) in which a foc us was given to automatic weld detection along the pipe and automatic interest areas segmentation, t his article focuses on the identification of pre-detected areas, using a complete pattern recognition process on totall y new data.
[1]
Leo Breiman,et al.
Classification and Regression Trees
,
1984
.
[2]
I. Jolliffe.
Principal Component Analysis
,
2002
.
[3]
Heng Tao Shen,et al.
Principal Component Analysis
,
2009,
Encyclopedia of Biometrics.
[4]
Leen-Kiat Soh,et al.
Texture analysis of SAR sea ice imagery using gray level co-occurrence matrices
,
1999,
IEEE Trans. Geosci. Remote. Sens..
[5]
Robert M. Haralick,et al.
Textural Features for Image Classification
,
1973,
IEEE Trans. Syst. Man Cybern..
[6]
N. Otsu.
A threshold selection method from gray level histograms
,
1979
.
[8]
Martin T. Hagan,et al.
Neural network design
,
1995
.
[9]
Automatic and Circumferential Oriented Detection (ACOD) for Pipeline Intelligent Inspection
,
2011
.
[10]
Leo Breiman,et al.
Random Forests
,
2001,
Machine Learning.
[11]
David A Clausi.
An analysis of co-occurrence texture statistics as a function of grey level quantization
,
2002
.
[12]
Rafael C. González,et al.
Digital image processing using MATLAB
,
2006
.