detection improvement of oil and gas pipelines with industrial radiography method by using image processing Alireza Karimian (Engineering faculty of Isfahan university, Isfahan,, Iran ) Sepideh Yazdani (Engineering faculty of Najaf Abad unit of Azad university, Isfahan,, Iran ) Amir Movafeghi (Department of NDT, Radiation Protection Technological Centre, INRA, Tehran, Iran ) Overview causes and Importance of its detection Radiography Test (advantages and limitations) The used digital image processing methods Results Conclusion is very often encountered in gas and oil pipes and caused by oxidation or chemical action of the corrosive materials Determining the extension and progress rate of internal and external corrosion and deposit provides useful information about the operational life and safety of the equipments and plants corrosion, erosion, deposits and blocking of pipes which might cause fire, leaks, reduced production or unpredictable and costly shutdowns due to repair and replacement 1
NDT inspections of pipelines Ultrasonic, magnetic and eddy current methods can be used in periodic inspections of pipelines Radiographic inspection is one of the most appreciated techniques among non- destructive testing methods due to the production of a film which acts as a fingerprint record. RT advantages Only radiographic technique assures inspection without costly removal of insulation material during operation of the plant The radiography can even be applied in high temperature environments Radiographic inspection produce a film which acts as a fingerprint record RT limitations Bad contrast and high un-sharpness of the radiograph quality, specially in the edge regions because of the scattered photons 2
To overcome the limitations, converting radiographs to a digital format and further digital image processing is the best method Objective Pipelines RT images improvement by using image processing techniques in MATLAB environment In this research work, After capturing the digital images by using the scanner, three methods were used to improve the image quality of pipelines RT images Methods Spatial technique Frequency domain technique Wavelet 3
Spatial technique steps 1. Convert image to gray scale 2. Apply the linear and nonlinear spatial filters (Laplacian( filter) 3. Apply histogram equalization (to improve the contrast) 4. Apply the median filter 5. Segmentation process Frequency domain steps 1. Convert image to gray scale 2. Obtain the Fourier transform of the image 3. Generate Low Pass Butterworth Filter 4. Multiply the transform by the filter 5. Obtain the Inverse Fourier transform 6. Apply histogram equalization Wavelet 1. Reduce the image noise 2. Making pseudo coloring plot 4
Spatial technique Original image of an oil pipeline on radiography film The converted image to Gray scale Gray scale double class image after laplacian filter Original Gray scaled image after Laplacian filter Subtraction of two images in pervious slides with histogram equalization and passing from a median filter 5
Segmentation and Put the threshold in the image Black area is probable corrosion and white region shows probability of deposit Comparison of original Gray scaled image with 2 last images Frequency domain technique example Original gray scaled image Fourier transform of image Applying the Butterworth low pass filter (n=1, d=100) 6
a) Convolution of two last images b) Inverse Fourier transform Of image (a) and applying Histogram equalization Example of using Wavelet method a) Original gray scaled image b) Reduce the image noise by Wavelet technique and making pseudo coloring plot Object with standard defects including holes and channels 7
Original gray scaled image Original gray scaled image Original gray scaled image Object image after using Spatial technique Object image after using Frequency domain technique Object image after using Wavelet technique 8
Imaged Area Gray scaled image of object Object with Gray scaled image of object After Spatial technique After Frequency domain technique Gray scaled image of object Object image after using Wavelet technique 9
Gray scaled image of object After Spatial technique Segmentation and Put the threshold in the image Gray scaled image of object After Frequency domain technique Gray scaled image of object Object image after using Wavelet technique 10
Original Gray scaled image Spatial technique Frequency Technique Wavelet Technique Conclusion Our results show by converting the radiography film images to digital images and using one of the above image processing methods, the corrosion and deposit with the dimensions about 1mm are detectable while in the same radiographic films these defects were not detectable Thank you for your attention 11