Title: Multi-sensor data fusion using geometric transformations for the nondestructive evaluation of gas transmission pipelines
1Multi-sensor data fusion using geometric
transformations for the nondestructive evaluation
of gas transmission pipelines
- by
- PJ Kulick
- Graduate Advisor Dr. Shreekanth Mandayam
- MS Final Oral Presentation
- August 29, 2003, 300 PM
2Outline
- Introduction
- Objectives and Scope of Thesis
- Background
- Approach
- Implementation Results
- Conclusions
3Gas Transmission Pipelines
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
- 280,000 miles
- 24 - 36 inch dia.
4In-Line Inspection
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
5Nondestructive Evaluation (NDE)
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
6Gas Transmission Pipeline Indications
- Benign
- T-sections
- Welds
- Valves
- Taps
- Straps
- Sleeves
- Transitions
- Anomalies
- Stress Corrosion Cracking
- Pitting
- Arching
- Mechanical Damage
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
7NDE using Multiple Inspection Modalities
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
8Data Fusion
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
9Data Fusion
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
10Objectives of This Thesis
- Develop data fusion techniques for the extraction
of redundant and complementary information - Validate techniques using simulated canonical
images - Validate techniques using laboratory NDE signals
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
11Expected Contributions
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
- A data fusion algorithm with the ability to
identify redundant and complementary information
present in multiple combinations of pairs of NDE
data sets. - i. e. (MFL-UT, MFL-Thermal, UT-Thermal)
12Nondestructive Evaluation of Gas Pipelines
0.2
0.0
0.6
0.4
Ultrasonic Testing
Magnetic Imaging
Virtual Reality
Data Fusion
Advanced Visualization
Acoustic Emission
- This research work is sponsored by
- US Department of Energy
- National Science Foundation
- ExxonMobil
Thermal Imaging
Digital Signal/Image Processing
Test Platforms
13Previous Work in Data Fusion
- Mathematical Theory
- Probability Theory
- Bayes Theorum
- Possibility Theory
- Fuzzy logic
- Belief Theory
- Dempster Shafer
- Improved DS Theories
- Transferable Belief Model
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
14Previous Work in Data Fusion
- Mathematical Transforms
- Discrete Fourier Transform (DFT)
- Discrete Cosine Transform (DCT)
- Wavelet based transforms
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
15Geometric Transformations
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
16Geometric Transformations
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
17Approach
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
Feature x1
Geometric Transformation
Redundant/ Complementary Information
OBJECT
Feature x2
g2(x2) T g1-1(x1, x2) h
homomorphic operator
18Approach
- Redundant Data Extraction
- ? Train RBF (homomorphic operator ? )
- g1(x1, x2) g2(x2) h1
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
19Approach
- Redundant Data Extraction
- ? Test RBF
- h1 x2 g1(x1, x2)
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
20Canonical Image Results
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
x2
x1
- 6 Images
- 4 Training
- 2 Test
- 20 x 20 pixels
- 20 x 20 DCT sent into network in vector form
Complementary
Redundant
21Canonical Image Results
- Simulation 1 Training Data Results
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
22Canonical Image Results
- Simulation 1 Test Data Results
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
23Canonical Image Results
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
x2
x1
- 6 Images
- 4 Training
- 2 Test
- 20 x 20 pixels
- 20 x 20 DCT fed into network in vector form
Complementary
Redundant
24Canonical Image Results
- Simulation 2 Training Data Results
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
25Canonical Image Results
- Simulation 2 Training Data Results
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
26Canonical Image Results
- Simulation 2 Test Data Results
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
27Experimental Setup
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
28Experimental Setup MFL
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
Pipe section
Hall probe
Probe mount
Current leads
Clamp
29Experimental SetupTangential MFL Images
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
30Experimental Setup UT
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
31Experimental SetupUT Time of Flight (TOF) Images
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
32Experimental Setup Thermal
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
33Experimental SetupThermal Phase Images
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
34What is Redundant and Complementary Information?
- We have defined this as follows
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
Defect Profile
Method 1 NDE Signature
Method 2 NDE Signature
Redundant Information
Complementary Information
35Experimental SetupTangential MFL Images
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
36Experimental SetupUT Time of Flight (TOF) Images
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
37Experimental SetupThermal Phase Images
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
38Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
Trial 1 UT-MFL UT-Thermal MFL-Thermal
Trial 2 UT-MFL UT-Thermal MFL-Thermal
Trial 3 UT-MFL UT-Thermal MFL-Thermal
39Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
40UT and MFL Data Fusion Results
Trial 1
41UT and MFL Data Fusion Results
Trial 1
42Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
43UT and MFL Data Fusion Results
Trial 2
44UT and MFL Data Fusion Results
Trial 2
45Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
46UT and MFL Data Fusion Results
Trial 3
47UT and MFL Data Fusion Results
Trial 3
48Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
Trial 1 UT-MFL UT-Thermal MFL-Thermal
Trial 2 UT-MFL UT-Thermal MFL-Thermal
Trial 3 UT-MFL UT-Thermal MFL-Thermal
49Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
50UT and Thermal Data Fusion Results
Trial 1
51UT and Thermal Data Fusion Results
Trial 1
52Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
53UT and Thermal Data Fusion Results
Trial 2
54UT and Thermal Data Fusion Results
Trial 2
55Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
56UT and Thermal Data Fusion Results
Trial 3
57UT and Thermal Data Fusion Results
Trial 3
58Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
Trial 1 UT-MFL UT-Thermal MFL-Thermal
Trial 2 UT-MFL UT-Thermal MFL-Thermal
Trial 3 UT-MFL UT-Thermal MFL-Thermal
59Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
60MFL and Thermal Data Fusion Results
Trial 1
61MFL and Thermal Data Fusion Results
Trial 1
62Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
63MFL and Thermal Data Fusion Results
Trial 2
64MFL and Thermal Data Fusion Results
Trial 2
65Data Fusion Trials
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
66MFL and Thermal Data Fusion Results
Trial 3
67MFL and Thermal Data Fusion Results
Trial 3
68Accomplishments
- Development of a generalized technique for fusing
data from two distinct observations of the same
object - Design of an algorithm that can extract redundant
and complementary information from two distinct
observations of the same object - Validation using simulated canonical images
- Validation using lab data representative of the
NDE of gas transmission pipelines
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
69Conclusions
- Algorithm is sufficiently general does not
specify which features are redundant or
complementary - Efficacy has been demonstrated by defining the
redundancy and complementarity of two NDE images
by correlating defect signature pixels with the
location, size and shape of the defect - Definition and approach are extremely accurate in
all instances of training data and sufficiently
accurate in all instances of test data - Information presented to the neural network is
distinct the matrices manipulated are
non-singular - The errors that occur during certain instances of
training and testing illustrate the need for a
large, more diverse data set - Data fusion of UT/MFL proved better then data
fusion of UT/Thermal or MFL/Thermal
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
70Directions for Future Work
- Enhancement of training and test data
- Explore variety of image preprocessing techniques
- Investigate various definitions of redundant and
complementary information - Test techniques robustness with noisy real-world
NDE signals - Adapt algorithm for heterogenous datasets
OUTLINE Introduction Objectives/
Scope Background Approach Implementation
Results Conclusions
71Acknowledgements
- U.S. Department of Energy, "A Data Fusion System
for the NondestructiveEvaluation of Non-Piggable
Pipes," DE-FC26-02NT41648 - ExxonMobil, "Development of an Acoustic Emission
Test Platform with a Biaxial Stress Loading
System," PERF 95-11 - Joseph Oagaro