International Journal of Trend in Scientific Research and Development (IJTSRD)

Volume 5 Issue 4, May-June 2021 Available Online: www.ijtsrd.com e-ISSN: 2456 – 6470

An In-depth Review on Bridge Crack Detection Approaches

Ketan Ovhal1, Ruturaj Lokhande1, Omkar Kamble1,
Prathamesh Nanaware1, Samarsingh Jadhav2
1Student, 2Professor,
1,2Department of Computer Engineering, PVPIT, Pune, Maharashtra, India

ABSTRACT How to cite this paper: Ketan Ovhal |

Bridges are mega structures that have been utilized and built for millennia. Ruturaj Lokhande | Omkar Kamble |
These structures are highly effective in achieving transportation and commute Prathamesh Nanaware | Samarsingh
between two highly inaccessible destinations easily. Bridges are also highly Jadhav "An In-depth Review on Bridge
effective in reducing traffic was by allowing the use of an alternate path for the Crack Detection Approaches" Published in
traffic flow to be resumed. This makes them highly versatile and extremely International Journal
effective in various scenarios. But as with any constructed structures, these of Trend in Scientific
bridges need to be evaluated for their structural integrity and surveyed for Research and
any flaws or cracks that have been emerged over time. This is usually done Development (ijtsrd),
manually by a civil engineer which is a time-consuming process and can also ISSN: 2456-6470,
introduce human error. Therefore to improve this procedure and number of Volume-5 | Issue-4,
related works have been analyzed extensively to achieve bridge crack June 2021, pp.957- IJTSRD42401
detection through image processing methodologies. An effective approach has 959, URL:
been envisioned through the use of convolutional neural networks and www.ijtsrd.com/papers/ijtsrd42401.pdf
decision tree techniques to achieve bridge crack detection which will be
further elaborated in the next edition of this research article. Copyright © 2021 by author (s) and
International Journal of Trend in Scientific
KEYWORDS: Convolutional Neural Networks, Decision Tree Research and Development Journal. This
is an Open Access article distributed
under the terms of
the Creative
Commons Attribution
License (CC BY 4.0)
(http://creativecommons.org/licenses/by/4.0)

I. INTRODUCTION:
Bridges are effective and highly useful structures that allow As the bridges are made in extreme environments they are
for transportation and computer between two points that subjected to environmental forces which can be slowly
are not easily accessible. Bridges allow foreign keys in damaging to the colossal bridge structure. The bridges need
computer and can effectively reduce the traffic problems by to be structurally perfect so that there is no mishap that can
creating another route for transport in the same amount of occur while using those bridges for transportation or
available space. This allows overlapping roads to effectively commute. Engineers are effectively performing regular
allow computers to go on their journey without any hiccups service on various bridges that are still standing today to
or interruptions by traffic signals or other traffic determine the amount of damage that is being done onto the
management devices. Bridges are also highly effective in bridge and provide any effective repairs before any mishap
bridging the gap between accessible entities wherein a road occurs. As bridges are used extensively there is a need to
cannot be constructed between the two for allowing effective maintain the structural performance in order to produce
transport. Bridges are highly useful and have been maximum performance from the bridges and achieve
constructed for millennia. effective improvements.
The bridges also can provide a path from one place to Usually, the process of bridge maintenance is done manually
another over extremely dangerous territories such as valleys where the engineers would physically examine the structure
or rivers for people to commute every day. These bridges are of the bridge along with various important landmarks such
useful and can be used to transport goods provide railway as beams columns, and other rigid structures for early signs
tracks or ease traffic conditions in a meaningful way. The of weakness or structural loss of integrity by looking out for
civil engineers design these bridges through an extensive cracks. This is a highly laborious process and it is extremely
survey of various different conditions of the environment time-consuming for the engineer to perform this survey
and produce the most efficient design to help implement in a which is highly essential for maintaining the integrity and
safe and structurally sound manner. There are various the structural reliability of the bridge intact. Therefore there
bridges across the world that are more than a hundred years is a need to improve this procedure through the
old and are still standing and being used regularly by the implementation of technology to achieve effective
people. The reason these bridges are still effectively useful is maintenance of the bridge structures. For this purpose, a
due to effective design methodologies and principles along number of researchers have been analyzed to achieve the
with highly effective maintenance. prescribed approach in the implementation of image

@ IJTSRD | Unique Paper ID – IJTSRD42401 | Volume – 5 | Issue – 4 | May-June 2021 Page 957
 International Journal of Trend in Scientific Research and Development (IJTSRD) @ www.ijtsrd.com eISSN: 2456-6470
processing approaches for identifying cracks in the bridge create a problem. [6] Thus the proposed paper implements a
structure. crack measurement system with the combination of a high
magnification image acquisition system and laser ranging
A large number of researchers have been analyzed for the
system. Thus the proposed paper implements an effective
purpose of achieving our goals of implementing image
approach that is efficient and convenient.
processing approaches for bridge crack analysis. These
approaches have been significant in achieving our approach H. Cho implements the detection of cracks in concretes by
for the implementation of an effective crack detection using an image-based methodology. In recent years there has
methodology that utilizes image normalization along with been lots of research made in the field of crack detection but
convolutional neural networks and the decision tree they have to face inaccurate extraction of edge pixels. [7]
approach. Thus in proposed paper consist of five steps such as crack
region search, crack width transform, hole filling, aspect
This literature survey paper dedicates section 2 for analysis
ratio filtering, and relative thresholding. Thus the
of past work as a literature survey, and finally, section 3
experiments were tested on synthetic and real images, which
concludes the paper with traces of future enhancement.
has resulted has consistency in measuring crack and greater
II. RELATED WORKS accuracy.
V.M. Jovanovic sequential-probability-ratio-test (SPRT) is
M.Lin elaborates the issue of the road such as Cracks. In
proposed and analyzed for Detector logic depends on the
recent years there is a lot of research and many funds
optimal. [1] This proposed technology is also compared with
invested for the automatic detection of cracks. Thus the
the heuristic sequential technique. Pseudo-noise waveform
proposed paper implements an accurate crack detection
is generated locally due to correlation received from the
method. [8] The method is divided into two parts one is
waveform. The simple “generalized” up-down counter logic
implemented using a hidden Markov random field model and
is approximated detector with devised. Thus this new
the adaptive line detector and the second is developed by
detector is compared with some heuristic sequential which
using the integrity conditional connection and by a
are analyzed in past.
quantitative description of crack regions.
T.Thesken introduces Continuous-phase modulation (CPM)
H Rowshande describes crack detection as one of the
for a relatively narrow spectrum. Is environment is coherent
promising topics of research in recent times. Depending on
by using an additive-white- Gaussian-noise (AWGN) for the
the low-speed walking stick method they have used an ACFM
optimal receiver. [2] They calculated the channel’s cutoff
sensor for classification and detection of RCF cracks. [9] The
rate, Ro. Just to determine the good a channel is when
CTSM algorithm is the existing system that has low-level
encoded. Thus to find the encoded data and to check the best
accuracy thus to overcome this shortcoming thus the
performance code. Thus by using random coding parameter
proposed paper implements a method for detecting multiple
Ro behavior of the channel with coding is analyzed on
RCF cracks. Thus the detected cracks can be maintained can
average.
be used to reduce inspection time.
Gary Garrabrant presents Additive White Gaussian Noise
Jingyi Li develops crack detection technology by using
(AWGN) channels for the non-coherent detection of
Science and technology. Thus the proposed paper develops a
orthogonal signals is implements trellis codes.[3] For a
crack detection system by using digital image processing.
communication system, an alternative coding technique is
[10] This system is divided into two parts as one part is the
developed by using non-coherent detection of the orthogonal
hardware part of the detection system and the second is the
signal set where the encoding can be expanded. To prove the
software part of the crack detection system which is
trellis coding is successful it is compared with a non-binary
extracted from the acquired images to detect the crack. Thus
block code.
in hardware image is collected from the car camera and
J.Zhang states beam bridges get cracks or girders due to signal processors. Thus the proposed paper is known as an
overweight freight vehicles, earthquakes. To detect the effective approach.
crackings and faults there is a lot of research made in recent
Yuwen Quan describes road rehabilitation road crack
years. [4] For fault detection purposes vertical and
detection as known as important for road crack detection.
longitudinal axial is observed by using the response signals.
[11] On gray histogram is proposed on the basis of Otsu
Thus to cracking fault detection by using traditional vertical
threshold crack detection. Thus the proposed methodology
vibration signal is very sensitive by measuring longitudinal
upgrades the accuracy of target extraction and keeps away
strain.
the prominent peaks. By using the road image processing of
S.Zhang presents object detection thus to detect objects of various crack types. There are the various steps Crack Image
present categories in the field of computer vision. Thus by Pre-Processing, Crack Image Segmentation. Thus the
using image recognition and video surveillance for accurate proposed paper results provide good consistency in
object detection there are various applications developed. detecting cracks.
[5] There various techniques used for crack detection such
W.Li elaborates china is known as one of the developing
as anchor-based detectors with the development of
nations in high-speed railways. There are various challenges
convolutional neural network (CNN). According to statistical
affected such as damage from external factors and aging of
features of an object, the proposed paper proposes an
their structures. [12] Thus the proposed paper implements
Adaptive Training Sample Selection (ATSS).
the crack detection method in a very complex environment.
Y.Cai describes one of the diseases that is commonly found The implement is used for histogram equalization just for
that is Cracks in tunnels, buildings, bridges, and other reducing uneven illumination for the detecting cracks and
structures. Cracks are often generated regularly or then Canny edge detection algorithm for adapting dual
irregularly it has to be repaired on time otherwise it will thresholds.

@ IJTSRD | Unique Paper ID – IJTSRD42401 | Volume – 5 | Issue – 4 | May-June 2021 Page 958
 International Journal of Trend in Scientific Research and Development (IJTSRD) @ www.ijtsrd.com eISSN: 2456-6470
J.Han explains in the concrete structure inspection field is Conference on Information Science and Technology,
known as one of the important tasks. There are much 2012, pp. 447-450, doi:
techniques implemented such as convolutional neural 10.1109/ICIST.2012.6221687.
networks thus they can crack results and blurred crack
[5] S. Zhang, C. Chi, Y. Yao, Z. Lei and S. Z. Li, "Bridging the
images [13]. Thus solutions for this introduces a multi-scale
Gap Between Anchor-Based and Anchor-Free
feature fusion technique for crack detection. Thus the
Detection via Adaptive Training Sample Selection,"
proposed architecture of CNN develops an encoder-decoder
2020 IEEE/CVF Conference on Computer Vision and
structure for image segmentation technique. Based on the
Pattern Recognition (CVPR), 2020, pp. 9756-9765,
quantitative measurements such as F-measure, recall, and
doi: 10.1109/CVPR42600.2020.00978.
precision.
[6] Y. Cai, X. Fu, Y. Shang and J. Shi, "Methods for Long-
III. CONCLUSION AND FUTURE SCOPE
Distance Crack Location and Detection of Concrete
An effective methodology for the purpose of bridge crack
Bridge Structures," 2018 IEEE 3rd International
detection through the use of image processing techniques
Conference on Image, Vision and Computing (ICIVC),
has been envisioned in this survey article. Bridges are highly
2018, pp. 576-580, doi:
useful structures that provide effective transportation and
10.1109/ICIVC.2018.8492764.
commute that would have not been possible without the
construction of these structures. These structures are [7] H. Cho, H. Yoon and J. Jung, "Image-Based Crack
usually massive in nature and allow for large trucks and Detection Using Crack Width Transform (CWT)
other transportation to be carried out effectively. These Algorithm," in IEEE Access, vol. 6, pp. 60100-60114,
mega structures are built by civil engineers through analysis 2018, doi: 10.1109/ACCESS.2018.2875889.
of various environmental conditions and a perfect plan for a
[8] M. Lin, R. Zhou, Q. Yan and X. Xu, "Automatic
bridge is generated and the bridge is built. These bridges
Pavement Crack Detection Using HMRF-EM
also need to be effectively maintained and surveyed for any
Algorithm," 2019 International Conference on
flaws or that have emerged over time due to this bridge is
Computer, Information and Telecommunication
being subject to harsh environmental conditions. To improve
Systems (CITS), 2019, pp. 1-5, doi:
these survey capabilities which is usually performed
10.1109/CITS.2019.8862067.
manually by engineer the utilization of image processing
techniques has been proposed in this survey article. This [9] H. Rowshandel, G. L. Nicholson, C. L. Davis and C.
server article has allowed for the creation of our approach Roberts, "A combined threshold and signature match
which implements convolutional neural networks along with method for the automatic detection of rail RCF cracks
a decision tree to achieve effective bridge crack detection. using an ACFM sensor," 6th IET Conference on
This approach will be further detailed in the upcoming Railway Condition Monitoring (RCM 2014), 2014, pp.
research article on this topic. 1-6, doi: 10.1049/cp.2014.0998.
REFERENCES [10] Jingyi Li, Ning Wang, Ying Liu and Yuemei Yang, "A
[1] V. M. Jovanovic, "Acquisition of frequency-hopping Study of Crack Detection Algorithm," 2015 Fifth
spread spectrum signals by sequential detection," International Conference on Instrumentation and
IEEE Military Communications Conference, 'Bridging Measurement, Computer, Communication and Control
the Gap. Interoperability, Survivability, Security', (IMCCC), 2015, pp. 1184-1187, doi:
1989, pp. 870-874 vol.3, doi: 10.1109/IMCCC.2015.254.
10.1109/MILCOM.1989.104044.
[11] Y. Quan, J. Sun, Y. Zhang and H. Zhang, "The Method of
[2] T. M. Thesken and E. Biglieri, "Cutoff rate for channel the Road Surface Crack Detection by the Improved
having precoding, continuous-phase modulation, and Otsu Threshold," 2019 IEEE International Conference
differential detection," IEEE Military Communications on Mechatronics and Automation (ICMA), 2019, pp.
Conference, 'Bridging the Gap. Interoperability, 1615-1620, doi: 10.1109/ICMA.2019.8816422.
Survivability, Security', 1989, pp. 764-768 vol.3, doi:
[12] W. Li, M. Zhang, Z. Shen, W. Hu and P. Li, "Track Crack
10.1109/MILCOM.1989.104026.
Detection Method in Complex Environment," 2018
[3] G. Garrabrant and J. Ehrenberg, "Trellis coded 11th International Symposium on Computational
modulation applied to noncoherent detection or Intelligence and Design (ISCID), 2018, pp. 356-359,
orthogonal signals," IEEE Military Communications doi: 10.1109/ISCID.2018.00087.
Conference, 'Bridging the Gap. Interoperability,
[13] J. H. Han, Y. C. Cho, H. G. Lee, H. S. Yang, W. J. Jeong and
Survivability, Security', 1989, pp. 774-778 vol.3, doi:
Y. S. Moon, "Crack Detection Method on Surface of
10.1109/MILCOM.1989.104028.
Tunnel Lining," 2019 34th International Technical
[4] J. Zhang, L. Li, X. Li and Z. Mu, "Crack detection of Conference on Circuits/Systems, Computers and
beam-like structures under environmental ground Communications (ITC-CSCC), 2019, pp. 1-3, doi:
motion excitation," 2012 IEEE International 10.1109/ITC-CSCC.2019.8793450.

@ IJTSRD | Unique Paper ID – IJTSRD42401 | Volume – 5 | Issue – 4 | May-June 2021 Page 959