Page 73 - Riem-Vol6_nº3
Basic HTML Version
418
IBRACON Structures and Materials Journal • 2013 • vol. 6 • nº 3
Monitoring of crack opening in masonry and concrete using mobile platform
A comparative test for different distances between the masonry structure
and the mobile phone camera (focal length) was made. The test allowed
finding that, for distances bellow 120 mm occurs the distortion of the im-
age seen on the mobile phone’s screen. This image distortion is noticed
by the user, because it shows the cracks with bigger ends compared to
the real dimension. There was also verified that with focal length above
280 mm occurs the absence of the visualization of the crack beginning
appearances, and bigger repeatability errors on measurements. In this
case, the optimized focal length, in order to avoid distortions or crack
identifications losses on the image, corresponds to 200 mm.
In Figure 4 it is illustrated the proposed acquisition system that
uses the mobile phone on the crack on masonry image acquisition
and processing.
Another characteristic of the proposed acquisition system is the
illumination system with a lamp directed to the crack’s region, as
illustrated in Figure 4(c). The implemented illumination system
has as objective avoiding luminosity interferences, for example,
the alteration of the environment light, shadows or reflexes during
the image capture process. The illumination system avoids bigger
errors on the repeatability of reading of the proposed method. In
Figure 4(c) it is possible to observe the details of the circular illumi-
nation device positioned at the image acquisition system.
The mobile phone can easily be removed from the acquisition sys-
tem in order to be utilized to its usual communicational applica-
tions, as can be seen in Figure 4(a).
4. Digital image processing
The proposed digital processing is accomplished at the spatial
plan of the image obtained from the structure on masonry or con-
crete. This processing acts directly on the pixels’ morphology of
image. Figure 5 illustrates the blocks diagram of the proposed im-
age processing.
The block 1 in Figure 5 represents the cracked structure on ma-
sonry or concrete image obtained by the acquisition system de-
scribed on Section 3. A crack on masonry image illustrated in Fig-
ure 6(a) is quantized on 256 gray levels.
On the block 2 of the block diagram from Figure 5, the image with
256 gray levels from Figure 6(a) is represented on the matrix form
by the function f(x, y), when x and y represent, respectively, the
coordinates for row and column in each pixel position.
The block 3 in Figure 5 normalizes the pixels intensity values (0-
255) for the interval between 0 and 1. This calculation is accom-
plished based on the equation (1) and the new values are stored
on the matrix g(x, y).
(1)
),(
)) ,(
max(
),(
yxf
yxf
yxg
Where
:
f(x, y) is the original intensity value(0≤ f(x, y) ≤ 255);
max(f(x, y)): is the highest value for the matrix f(x, y) and g(x, y) is
the resultant matrix with normalized values (0≤ g(x, y) ≤ 1).
The function of the block 4 in Figure 5 is to implement the bina-
rization process of the matrix g(x, y). In this case, was applied the
segmentation technique for image thresholding as described on
[24] and mathematically described by the equation (2).
(2)
L yxg
L yxg
yxh
) , ( se 0
,
) , ( se 1
) ,(
On the proposed method the image’s threshold L on equation (2)
is adjusted by the user in the beginning of the crack detection pro-
cess. The value is determined aiming to identify the beginning of
the crack’s appearance in reason of the tonality on the structure on
masonry or concrete. The surface of the structure that will be moni-
tored must be prepared, and will be described on the next section.
In the obtained results, the threshold value was equal 130 on all of
the proposed method monitoring.
Figure 6(b) illustrates the binary image resultant of the process accom-
plished by the block 4. The measuring of the interest pixels quantity is
accomplished by the block 5 (Figure 5) that implements the equation (3).
(3)
qmpn
N
n
M
m
Q
1
0
1
0
Figure 5 – Block Diagram of the image
digital process of the proposed method
