397 ACI Structural Journal/May-June 2015
ACI STRUCTURAL JOURNAL TECHNICAL PAPER
Damage and loss assessment of reinforced concrete elements are
based in part on the length, width, and areal density of cracks.
Crack information is traditionally collected using a crack width
card and transferred to drawing sheets, which is both approximate
and labor-intensive. An automated procedure involving digital
image processing was developed and deployed to collect and
process crack data. The procedure was developed and validated
using data from the cyclic testing of nine reinforced concrete shear
walls of varying aspect and reinforcement ratios.
Keywords: automated crack detection and measurement; damage assess-
ment; imaging.
INTRODUCTION
The length, width, and areal density of cracks are used
for damage assessment of reinforced concrete compo-
nents in the laboratory and in the ield. Damage to rein-
forced concrete components in buildings and infrastructure
after earthquake shaking is often inferred using maximum
residual crack widths. Strategies for post-earthquake repair
are often based on this information. Cracking of reinforced
concrete components during laboratory testing is routinely
documented (along with other information) to enable a
reconciliation of loss of strength and stiffness with observed
damage. Information on the cracking of low aspect-ratio
reinforced concrete shear walls
1
has been used to generate
fragility functions and consequence functions that enable an
estimate of the repair cost per unit area of wall.
Engineers have traditionally identiied cracks visually in
reinforced concrete components, measuring their widths
using a crack-width card and then transferring that data
to drawing sheets. Crack width is measured either at user-
determined locations along the length of a crack, or a
maximum crack width is reported. Cracks are assigned a
width equal to one of the marks on the gauge (for example,
0.005 in., 0.016 in., or 0.06 in.; refer to Fig. 1). Widths of
cracks between the marks on a gauge can only be estimated.
The process is laborious and approximate because uncer-
tainty is introduced in the measurement of the crack width,
the use of few measurement locations, and the transfer of
information to drawing sheets.
Imaging tools provide a means to improve the process
of measuring and documenting cracks and their widths,
and to substantially improve the quality and accuracy of
the results. Noncontact identiication and measurements of
cracks enables data to be gathered from large-size labora-
tory tests at instances of peak loading and deformation in a
safer and more reliable manner. Gathering information on
the length, width, and areal density of cracks at instances of
peak deformation and zero loading enables estimates to be
made of structural damage to components, which are better
correlated with peak transient crack widths and lengths than
residual crack widths and lengths.
An automated process for detecting cracks, and measuring
their widths and lengths is presented in this paper. Data
from the tests of nine large-size, low aspect-ratio rein-
forced concrete shear walls are used to develop and vali-
date the algorithms and computer codes. The algorithms,
source code, and documentation will be uploaded to
NEEShub (www.neeshub.org) for use by researchers and
forensic engineers.
RESEARCH SIGNIFICANCE
An automated, noncontact procedure is proposed for the
estimation of crack width, length, and areal density. The
proposed procedure, which could be used in the ield and the
laboratory, is safer, faster, and more accurate than the tradi-
tional method of measuring lengths and widths of cracks by
hand using crack gauges and tapes and then transferring that
information to drawing sheets.
LITERATURE REVIEW
Image-based crack-detection algorithms have been used
primarily to detect cracks on pavement surfaces. Most
crack-detection algorithms rely on edge detectors to locate
Title No. 112-S32
Automated Detection and Measurement of Cracks in
Reinforced Concrete Components
by Jonathan P. Rivera, Goran Josipovic, Emma Lejeune, Bismarck N. Luna, and Andrew S. Whittaker
ACI Structural Journal, V. 112, No. 3, May-June 2015.
MS No. S-2014-084.R1, doi: 10.14359/51687424, received August 18, 2014, and
reviewed under Institute publication policies. Copyright © 2015, American Concrete
Institute. All rights reserved, including the making of copies unless permission is
obtained from the copyright proprietors. Pertinent discussion including author’s
closure, if any, will be published ten months from this journal’s date if the discussion
is received within four months of the paper’s print publication.
Fig. 1—Crack gauge.
11
(Note: 1 in. = 25.4 mm.)