Application of Smart Sequence Reconsruction Technique to PVT
Images of Composite Car Body Components
MAHMOUD Z. ISKANDARANI1 AND NIDAL F. SHILBAYEH2
1
Department of Computer Engineering
2
Department of Computer Science
Applied Science University
1
P.O.BOX 911597, Post Code: 11191, Amman, 2P.O.BOX 41, Post Code: 11931, Amman
JORDAN
Abstract: - An effective NDT (Non-Destructive Testing) image analysis technique for detecting materials
damage and defects existance has been developed successfully and applied to PVT images of car wishbone.
The developed technique is based on converting an image to its equivalent pixel values and then applying
smart reconstruction algorithm to the converted image such that the presence of damage in the composite
strucure and its extent can be easily verified.
Key-Words: - Non-Destructive Testing, Defect Detection, Thermography, Image Reconstruction, Intelligent
Systems, PVT.
1 Introduction
The layered composites are presently the most
widespread advanced materials in use. Among
them, fiber reinforced composites with polymeric
matrices (FRP or laminates) and polymeric
sandwich materials, with thin laminate faces and
foam or impregnated cores, are in real progress.
The structural design and maintenance of
composite structures involving these materials need
comprehensive evaluation and characterization of
mechanical properties and behavior under different
loading conditions, in both undamaged and
damaged state. [1,2,3,4,5,6].
The marked inhomogeneity and anisotropy of
these materials makes them vulnerable to a variety
of damages. For this reason, reliable composite
structures need adequate NDT/NDE methods along
the maintenance activities and knowledge of
residual strength/stiffness or service life estimation
linked to certain damage patterns. In the end,
development of damage tolerant materials may be
considered a goal towards further increasing the
attractiveness of composite materials in building
high tech reliable products.
Many NDT methods were proposed and are in
use for evaluating the structural integrity of
composites, from simple visual inspection to laser
shearography, with various sensitivity, versatility
and affordability.
In order to have a versatile NDT inspection
method, ready to be used in industrial applications,
not merely for laboratory research, the prime
requirement is to need access on only one side
using Pulse Video Thermography (PVT) assisted
by computer and intelligent software specifically
designed for this purpose.
In this paper a new technique in detecting and
analyzing the presence of damage in a car
composite structure is presented. A wishbone
component is used as a testing example to verify
the validity of the technique. The wishbone is
impact damaged before testing.
2 Background
IRT for NDE is aimed at the discovery of
subsurface features (such as subsurface thermal
properties,
presence
of
subsurface
anomalies/defects), thanks to relevant temperature
differences observed on the surface with an infrared
(IR) camera. Figure 1 illustrates the general
concept. IRT is deployed along two schemes,
passive and active. The passive scheme tests
materials and structures which are naturally at
different (often higher) temperature than ambient
while in the case of the active scheme, an external
stimulus is necessary to induce relevant thermal
contrasts (which are not available otherwise, e.g.
specimen at uniform temperature prior to testing)
[7,8,9,10,11].
Each NDE technique has its own strengths and
weaknesses. In the case of thermography these are
as follows:
I. Fast inspection rate.
II. No contact.
III. Safety.
IV. Results relatively easy to interpret
V. Wide range of applications.
On the other hand, difficulties are as follow:
I. Difficulty to deposit uniformly a large amount
of energy in short period of time over a large
surface.
II. Effects of thermal losses (convective,
radiative, conductive) perturbating thermal
contrasts.
III. Cost of the equipment (IR camera, thermal
stimulation units for active thermography).
IV. Capability to detect only entities (subsurface
defects) resulting in a measurable change of
thermal properties.
V. Ability to inspect a limited thickness of
material under the surface (thermography is a
‘boundary technique).
VI. Emissivity problems
2.1
Pulse Video Thermography
Basically, pulsed thermography (PVT) consists of
briefly heat the specimen and then record the
temperature decay curve. Qualitatively, the
phenomenon is as follow. The material changes
rapidly after the initial thermal pulse because the
thermal front propagates, by diffusion, under the
surface and also because of radiation and
convection losses. The presence of a defect reduces
the diffusion rate so that when observing the
surface temperature, defects appear as areas of
different temperatures with respect to surrounding
sound areas once the thermal front has reached
them. Consequently, deeper defects will be
observed later and with a reduced contrast.
In fact, the observation time t is function (in a
first approximation) of the squared of the depth z
and the loss of thermal contrast c is proportional to
the cube of the depth:
t
z2
1
and c 

z3
Where α
is the thermal diffusivity of the
material.
These relations indicate two limitations of the
IRT: observable defects will generally be shallow
and the thermal contrasts will be weak. An
empirical rule of thumb says that the radius of the
smallest detectable defect should be at least one to
two times larger than its depth under the surface.
This rule is valid for homogeneous isotropic
material. In case of anisotropy it is more
constrained.
Various deployments are possible:
 point inspection (example: laser or focused
light beam heating),
 line inspection (example: heating with line
lamps, heated wire, line of air jets (cool or
hot),scanning laser),
 surface inspection (example: heating using
lamps, flash lamps, scanning laser); either in
reflection (thermal source and detector located
on the same side of the inspected component)
or in transmission (heating source and detector
located on each side of the component). If the
temperature of the part to inspect is already
higher than ambient temperature due to the
manufacturing process for instance, it might be
convenient to make use of a cold thermal
source such as a line of air jets. Obviously, a
thermal front propagates the same way whether
being hot or cold: what is important is the
temperature differential between the thermal
source and the specimen. Another advantage of
a cold thermal source is that it does not induce
spurious thermal reflections into the IR camera
as in the case of a hot thermal source.
Knowledge of the evolution of thermal contrast
above the defect in conjunction with equations
derived from inverse heat transfer modeling allows
to retrieve defect parameters such as depth,
diameter, thermal resistance. A common definition
of the thermal contrast C is:
C (t ) 
Ti (t )  Ti (t0 )
Ts (t )  Ts (t0 )
Where T is the temperature signal, t is the time
variable, subscripts i and s refer respectively to
over a suspected defective location (that is in fact
any pixel in the image) and over sound areas
respectively. C is computed with respect to before
heating temperature distribution at time t0 (to
suppress the adverse contributions from the
surrounding environment) and normalized by the
behavior of a sound area so that a unit value is
obtained over a non defect area. Such kind of
analysis is common in the automotive industry.
Other common applications of the active PVT
scheme are in quantitative subsurface defect
assessment
(cracks,
delaminations,
impact
damages, disbondings, moisture), thermophysical
property evaluation; in all kind of industries [12,
13, 14, 15, 16, 17].
3 System description
3.1 System Design
Fig.1 shows the used testing system. The system
consists of the normal PVT system with the
addition of the innovated sampling and sequence
reconstruction technique which is run under smart
environment that provides the overall interpretation
Fig.1 Image Acquisition and Sequence Reconstruction System
Fig.2 Block diagram of the Smart Environment
Fig.1 illustrates the process that the image goes
through from the capturing stage into interpretation.
Fig.2 illustartes the system used to learn about
diiferent types of damages for future predictions.
After capturing of the IRT image, it is sampled, split
or sliced into sequences and statistically regrouped
in comparison to intensity thresholds.
Fig.3 shows the capured images for a car
wishbone over intervals of time, where tables 1 to 5
illustrates the reconstructed image sequences for
each captured image at a specific time interval.
From the images and tables shown, we deduce
that using sequence reconstruction can be very
effective in determining the real existance of
damage. Following the increase in threshold value
from table 1 through table 5, which correspond to
time interval 0 seconds to 574 seconds, we notice
that pixel levels did not return back to its initial
value at the end interval indicating existance of
damage. This can be analyzed as follows:
I. t = 0 S: The first image in fig.3(a) shows the
initial application of a heat pulse to the
wishbone, with tables 1 through 3 (Low
Thresholds) indicating high statistical pixel
gathering as a result of low temperature
absorption at this time, while tables 4 through 5
indicate no heat absortption at High
Thresholds.
II. t = 94 S: An increase in the statistical pixel
regrouping is noticed in certain parts of the
tables as shown in fig.2(b).This indicates more
heat wave absorption as the time progressed
and an initial indication of the presence of
damage as heat waves take longer to diffuse
across damaged areas. This is in agrrement
with theory and literature.
III. t = 229 S: Start of heat decay and temperature
recovery process, except for the severly
damaged areas, which will stay at above a
statistically normal level as shown in fig.3(c).
IV. t = 574 S: Enough time for most of the
wishbone structure to recover, except for the
critically damaged area shown as large pixel
values in the tables as shown in fig.3(d).
Threshold
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
t = 0S
t = 94S
0
1
t = 229 S t = 574S
0
1
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4081
4148
4065
4000
9262
9347
13578
14362
2792
2807
4254
4653
168
152
20
22
102
103
6
4
60
64
1
1
49
31
1
0
31
15
1
1
12
5
3
0
11
93
17
68
1202
1255
1154
1206
15954
15878
26340
27149
6781
6457
11553
11224
468
355
140
83
252
155
3
6
169
93
0
3
130
58
1
3
35
43
0
0
32
14
2
1
16
36
7
26
49
51
17
11
1802
1945
460
330
889
1086
533
247
72
78
109
28
18
22
10
2
4
9
2
1
4
2
0
0
2
3
1
0
1
1
0
1
1
1
0
2
33
27
8
2
1501
1523
16
8
827
895
36
13
116
145
25
4
Table 1 First Level Sequence Reconstruction
Threshold
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
t = 0S
t = 94S
0
0
t = 229 S t = 574S
0
0
21
46
12
1
15
31
5
0
4
4
0
0
3
1
0
0
0
0
0
0
1
0
3
1
28
24
3
4
1183
1284
12
4
823
881
6
4
166
208
11
3
54
71
3
1
22
46
1
0
19
17
0
1
5
9
0
1
0
0
0
0
1
1
1
1
134
119
7
96
3944
4269
81
99
3743
3774
17
14
1470
1380
5
1
1006
963
2
1
502
521
1
1
286
234
0
0
158
126
0
2
65
46
1
0
126
127
102
179
426
424
494
484
4778
4588
7490
7648
1868
1995
2962
3094
176
139
22
13
131
99
11
6
100
73
2
1
86
38
2
1
64
8
1
1
23
12
0
0
19
33
2
31
272
257
306
293
10107
9929
17489
17810
4581
4778
8351
7990
432
375
172
100
280
255
13
20
248
130
7
4
197
84
1
2
125
66
1
0
63
47
1
2
44
31
2
18
26
28
7
4
1229
1333
405
265
708
709
609
258
72
74
291
128
Table 2 Second Level Sequence Reconstruction
Threshold
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
t = 0S
t = 94S
19
27
t = 229 S t = 574S
115
39
5
13
16
15
3
4
4
1
1
1
1
1
1
0
0
1
0
1
0
0
19
21
3
2
987
993
9
2
601
753
55
10
143
187
114
24
59
82
115
11
24
42
33
4
9
10
8
1
1
5
0
0
1
1
0
1
0
2
2
1
15
12
2
0
827
845
3
3
641
638
8
9
222
257
19
10
169
189
20
7
67
96
25
2
24
48
4
0
8
18
0
0
1
3
0
0
0
1
0
0
46
45
4
8
2741
2902
36
19
2724
2783
8
7
1360
1363
18
6
1253
1207
13
13
1085
1113
8
2
629
655
11
3
415
360
5
2
223
192
1
0
122
70
21
16
4
0
4
3
5
3
7
4
1
1
0
0
1
0
0
0
5
6
6
5
8
5
3
9
8
10
4
5
7
7
6
0
7
2
0
0
6
9
1
1
0
1
0
1
0
1
6
5
2
1
2
1
0
0
0
0
Table 3 Third Level Sequence Reconstruction
Threshold
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
t = 0S
t = 94S
11
11
t = 229 S t = 574S
2
9
18
17
5
5
25
22
8
6
17
9
3
6
16
15
2
1
30
28
0
1
0
1
0
0
0
0
0
1
0
2
0
1
1
3
2
1
2
0
6
3
0
3
6
1
3
2
2
3
0
0
3
3
1
0
1
2
2
2
0
2
1
1
0
0
2
0
0
1
2
0
0
0
2
4
5
0
5
5
15
0
1
3
8
2
1
2
5
1
1
2
0
1
0
1
1
3
1
0
0
2
3
0
1
0
0
1
1
0
0
1
0
0
3
3
3
0
9
11
8
4
9
5
6
2
6
2
6
4
2
5
2
3
0
2
4
0
0
1
2
2
0
0
0
0
3
5
0
1
9
7
0
1
10
15
6
3
4
40
35
6
132
113
7
9
139
130
13
15
82
85
9
16
58
69
6
7
125
117
11
9
70
24
131
13
16
37
103
18
1
1
2
2
0
0
0
1
Table 4 Third Level Sequence Reconstruction
Threshold
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
t = 0S
t = 94S
1
1
t = 229 S t = 574S
0
1
1
1
0
1
4
3
2
0
3
0
3
2
2
3
2
1
227
151
71
143
25
9
30
4
15
38
38
9
2
3
2
2
0
0
2
0
0
0
0
0
3
1
0
1
3
5
3
4
4
5
5
4
9
7
6
7
89
79
54
86
3
2
1
2
1
4
2
2
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
1
0
4
1
2
1
3
1
4
6
0
9
35
36
30
33
0
1
0
1
1
0
1
0
2
0
1
1
0
0
1
0
0
0
1
1
1
1
1
0
0
0
2
0
2
1
0
3
3
5
6
4
49
31
29
35
1
1
3
0
7
6
0
1
0
2
0
0
0
1
0
0
0
0
0
0
0
1
1
0
3
0
1
0
4
1
3
1
2
0
2
2
24
27
26
34
5
30
28
31
29
107
27
73
5
17
1
1
2
2
1
1
2
3
12
2
9
10
11
6
33
30
21
11
46
35
29
22
1172
953
1128
1014
Table 5 Fifth Level Sequence Reconstruction
Fig.3 Image Obtained Through PVT
4 Conclusion and Future Work
The emerging method of Pulsed Video
Thermography (PVT) can be used as a viable
Alternative or screening procedure for the
traditional NDT methods. It can detect the same
type of defects but does not require two sided
access, typically has a fast area scan rate, and can
be non-contact. A test specimen is pulsed with a
powerful externally applied heat source to create a
traveling steep temperature gradient within the test
specimen. Flaws or irregularities alter the flow of
heat, producing temperature contrasts at the surface
that are video captured via computer.
Thermography techniques can then be applied to
the surface image to characterize the flaw and
predict performance capability. Possible limitations
of PVT include the requirement of uniform surface
heating and the need for greater computational
speed/memory when defects are multi-dimensional
[18,19,20,21,22].
The development of a PVT computer work
station using off-the-shelf components would
have the benefits of low cost, ease of use,
portability, and flexibility of application.
In this paper a new algorithm and techniques is
provided and validated. This innovated technique
is capable of fast analysis of caputred PVT images
and other image formats, with smart engine based
on feature extraction and Neural Networks adding
the advantage of damage predicition. Further
development of the system is possible to improve
its accuracy and widen its range of application
[23,24,25].
References
[ 1 ] A. Dillenz, T. Zweschper, G. Riegert, and G.
usse, Progress in phase angle thermography,
Review of Scientific Instruments, Vol 74(1),
2003, pp. 417-419.
[ 2 ] L. D. Favro, Xiaoyan Han, Zhong Ouyang,
Gang Sun, Hua Sui, and R. L. Thomas,
Infrared imaging of defects heated by a sonic
pulse, Rev. Sci. Inst. 71, 6, 2000, S. 24182421.
[ 3 ] F. Galmiche, S. Vallerand, X. Maldague,
Pulsed Phase Thermography with the
Wavelet Transform, AIP Conference
Proceedings, Vol 509(1), 2000, pp. 609-616.
[ 4 ] X. Maldague, J.P. Couturier, D. Wu, A.
Salerno "Advances in Pulse Phase
Thermogra phy, QIRT-96 (Quantitative
Infrared Thermography), Eurotherm Seminar
50, Stuttgart (Germany/Allemagne), 1996,
pp 2-5.
[ 5 ] Th Zweschper, A. Dillenz, G. Riegert, D.
Scherling and G. Busse, Ultrasound excited
thermography using frequency modulated
elastic waves, Insight Vol 45 No 3, 2003.
[ 6 ] R. OSIANDER R. , J.W.M SPICER., Timeresolved infrared radiometry with step
heating. A review, Applied Physics Laboratory,
Vol. 37 , No 8 , 1998 , pp. 680 – 692.
[ 7 ] J.W.M Spicer., D.W. Wilson, R. Osiander ,
J. Thomas J., B.O. Oni, Evaluation of high
thermal conductivity graphite fibers for
thermal
management
in
electronics
applications, in Thermosense XXI, Proc.
SPIE, R.N. Wurzbach, D.D. Burleigh eds.,
1999, 3700: 40-47.
[ 8 ] G. Busse , D. Wu , W. Karpen , Thermal
wave imaging with phase sensitive
modulated thermography, J. Appl. Phys., Vol
71(8), 1992, pp. 3962-3965.
[ 9 ] G. Busse, Nondestructive evaluation of
polymer materials, NDT &E Int’l, Vol
27(5),1994, pp. 253-262.
[10 ] D. Wu, , A. Salerno , U. Malter , R. Aoki , R.
Kochendِrfer , P. K‫ن‬chele P, K. Woithe , K.
Pfister , G. Busse , Inspection of aircraft
structural
components
using
lockinthermography,
QIRT-96
(Quantitative
Infrared
Thermography),
Eurotherm
Seminar 50, D. Balageas, G.Busse, C.
Carlomagno eds., Edizioni ETS (Pisa, Italy) ,
1996, pp 251-256.
[11] Balageas D., Levesque P., “EMIR: A
photothermal tool for electromagnetic
phenomena characterization,” Rev. Gén.
Therm., 1998, 37: 725-739.
[12] L. Tenek , E. Henneke , Flaw dynamics
and vibro-thermographic thermoelastic
NDE of advanced composite materials,
in Thermosense XIII, Proc. SPIE, G. S.
Baird ed., 1467, 1991, pp. 252-263.
[13] R. Dinwiddie, P. Blau , Time-Resolved
Tribo-Thermography, in Thermosense XXI,
Proc. SPIE, R.N. Wurzbach, D.D. Burleigh
eds., 3700, 1999, pp. 358-368.
[14] J. Rantala , D. Wu , G. Busse , Amplitude
modulated lock-in vibrothermography for
NDE of polymers and composites, in
Research in NDE, Vol 7, 1996, pp. 215-228.
[15] A. Salerno, D. Wu, G. Busse, J. Rantala,
Thermographic inspection with ultrasonic
excitation, in Proc. Of Rev. Progresses in
Quantitat. NDE, D.O. Thompson, D.E.
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
Chimenti eds., NY: Plenum press, 16A, 1996,
pp.345-352.
X. Maldague, S. Marinetti, Pulse Phase
Infrared Thermography, J. Appl. Phys, Vol
79(5), 1996, pp. 2694-2698.
D. Prabhu, P. Howell, H. Syed, W. Winfree,
Application artificial neural networks to
thermal detection of disbonds, 1992, pp.
1331-13382 .
H. Trétout , D. David , J. Marin, M.
Dessendre M., An evaluation of artificial
neural networks applied to infrared
thermography inspection of composite
aerospace structures, Review of Progress in
Quantitative NDE, D.O. Thompson, D.E.
Chimenti eds, 14, 1995, pp. 827-834.
M. Hagan, H. Demuth , M. Beale , Neural
networks
design,
PWS
publishing
company,1996.
Santey M.B., Almond D. P., “An artificial
neural network interpreter for transient
thermography
image data”, NDT & E Int., 30[5]: 291-295,
1997.
P. Bison, S. Marinetti, G. Manduchi, E.
Grinzato, Improvement of neural networks
performances in thermal NDE, American
Soc. of Non Destructive Testing Press, Vol.
3, 1998, pp. 221-227,.
X. Maldague, Y. Largouët, Depth study in
pulsed phase thermography using neural
networks:Modeling, noise, experiments,
Revue Générale de Thermique, Vol 37(8),
1998, pp. 704-708.
B. Foucher , Infrared machine vision, in
Thermosense XXI, Proc. SPIE, R.N.
Wurzbach, D.D. Burleigh eds., 3700, 1999,
pp. 210-213.
G. d’Ambrosio, R. Massa , M. Migliore,
Microwave excitation for thermographic
NDE: An experimental study and some
theoretical evaluations Materials Evaluation,
Vol 53(4), 1995, pp. 502-508.
T. Sakagami, S. Kubo , Proposal of a new
thermographical
nondestructive
testing
technique using microwave heating, in
Thermosense XXI, Proc. SPIE, R.N.
Wurzbach, D.D. Burleigh eds., 3700, 1999,
pp. 99-103.