The use of images, whether in routine maintenance, or postearthquake reconnaissance, has quickly become the preferred approach to record and archive the exterior damage of existing infrastructure. Postsurvey analysis of these images, coupled with careful record keeping, provide invaluable data regarding the health of a structure. However, often significant amounts of data are obtained, especially for large structures, such as bridges. Therefore an automated procedure, which reliably and robustly reports on damage observed from these images, with minimal human intervention, is desirable. To this end, in this work, we present a statistical-based method for conducting image analysis, specifically for the purpose of evaluating concrete damage (cracks, spalling, etc.). We illustrate the derivation of the method, which is grounded in Bayesian decision theory and subsequently present results of the analysis of images with discrete cracks to illustrate its promise.

Unusually severe structural damage was reported during the 2010 M 7.0 Haiti earthquake in the vicinity of Hotel Montana, located on top of a ridge in the district of Pétionville. Prompted by the observations, U.S. Geological Survey seismic stations were deployed, and aftershock recordings indicated ground‐motion amplification on the top of the hill compared to adjacent stations on reference site conditions. The presence of topographic relief has been shown to significantly aggravate the consequences of strong ground motion during past events, and topographic effects were brought forward to explain the observations. In this paper, we test the hypothesis of topographic amplification as the dominant factor that contributed to the damage concentration in the vicinity of Hotel Montana. We initially conduct numerical simulations of the ridge seismic response assuming elastic homogeneous site conditions, and show that numerical predictions of topographic amplification disagree with the field data both in amplitude and in frequency. Conversely, while 1D ground‐response analyses for the site conditions at the hilltop predict amplification in the same frequency range as the field data, they significantly underestimate the recorded amplitude. We then conduct numerical simulations of the foothill ridge response to seismic motion while accounting for soil layering, and qualitatively demonstrate that the recorded amplification is most likely attributed to coupled site–topographic amplification effects, namely to seismic waves trapped in the soft soil layers of the near surface, amplified as a consequence of reverberations, and further modified due to diffraction and scattering upon incidence on the irregular ground surface. Parametric investigations of the topography–soil amplification coupling effects are then conducted, and our results show that when accounting for a hypothetical soil–bedrock interface at 100 m depth, predictions are in excellent agreement with the observed motion.

Although crack inspection is a routine practice in civil infrastructure management (especially for highway bridge structures), it is time‐consuming and safety‐concerning to trained engineers and costly to the stakeholders. To automate this in the near future, the algorithmic challenge at the onset is to detect and localize cracks in imagery data with complex scenes. The rise of deep learning (DL) sheds light on overcoming this challenge through learning from imagery big data. However, how to exploit DL techniques is yet to be fully explored. One primary component of practical crack inspection is that it is not merely detection via visual recognition. To evaluate the potential risk of structural failure, it entails quantitative characterization, which usually includes crack width measurement. To further facilitate the automation of machine‐vision‐based concrete crack inspection, this article proposes a DL‐enabled quantitative crack width measurement method. In the detection and mapping phase, dual‐scale convolutional neural networks are designed to detect cracks in complex scene images with validated high accuracy. Subsequently, a novel crack width estimation method based on the use of Zernike moment operator is further developed for thin cracks. The experimental results based on a laboratory loading test agree well with the direct measurements, which substantiates the effectiveness of the proposed method for quantitative crack detection.