Soroush Mokhtari

Improvement of Crack-Detection Accuracy Using a Novel Crack Defragmentation Technique in Image-Based Road Assessment

AbstractA common problem of crack-extraction algorithms is that extracted crack image components are usually fragmented in their crack paths. A novel crack-defragmentation technique, MorphLink-C, is proposed to connect crack fragments for road pavement. It consists of two subprocesses, including fragment grouping using the dilation transform and fragment connection using the thinning transform. The proposed fragment connection technique is self-adaptive for different crack types, without involving time-consuming computations of crack orientation, length, and intensity. The proposed MorphLink-C is evaluated using realistic flexible pavement images collected by the Florida Department of Transportation (FDOT). Statistical hypothesis tests are conducted to analyze false positive and negative errors in crack/no-crack classification using an artificial neural network (ANN) classifier associated with feature subset selection methods. The results show that MorphLink-C improves crack-detection accuracy and reduces...

Multi-Criteria Decision Making under Uncertainty: Application to California's Sacramento-San Joaquin Delta Problem

Multi-criteria decision making problems are often associated with trade-offs between performances of the available alternative solutions under the considered decision making criteria. These problems become more complex when uncertainty is associated. This paper applies a Monte-Carlo multi-criteria decision making method for solving the California’s Sacramento-San Joaquin Delta problem in which the best alternative for exporting water from the Delta should be selected based on two main criteria, namely the economic cost and environmental performance. To deal with the uncertainty involved, the stochastic decision making problem is converted into numerous deterministic problems through a Monte-Carlo selection. Each deterministic problem is then solved using a range of multi-criteria decision making methods. The overall winning probabilities of the alternatives under each method are calculated and compared with the results obtained in previous studies which used conflict resolution methods, namely, non-cooperative game theory, fall-back bargaining, and social choice rules to suggest optimal solutions for the Delta problem.

A Multi-Participant, Multi-Criteria Analysis of Energy Supply Sources for Fairbanks, Alaska

The selection of a future energy source for Fairbanks, Alaska, is a multi-criteria, multi-decision maker (MCMDM) problem as it involves a range of stakeholders who must consider economic, sociopolitical, and environmental criteria in deciding the best project. The primary motivation for the new energy project is to provide an additional affordable heating and electric source to local residents. Proposed projects range from liquid-natural gas pipelines to hydropower and differ greatly in development costs, environmental impacts, and political support. Stakeholder interests vary from local and state government officials, to local and international business developers, and to residents and environmentalists. Traditionally, water and resource MCMDM problems have been simplified to analyze a single decision maker (DM) for multiple criteria; this work defines model criteria at different levels based on input from stakeholder representatives through a collaborative process. Model inputs can be ordinal when cardinal information is unavailable, thereby increasing the models flexibility for a wide range of source data. The model employs a range of social choice, fall back bargaining, and MCDM to solve the problem. Uncertainty in the model is characterized by a Monte Carlo analysis, which measures sensitivity of the solution from the range of inputs provided by stakeholder data. Given the economic and social components included in this MCMDM analysis, characterizing the uncertainty associated with each outcome is crucial for policy interpretation. This work provides a new application for MCMDM problems combining a range of social choice and game theoretic methods with a rigorous sensitivity analysis to inform decision makers about the most feasible and stable alternatives.

Crack Recognition and Segmentation Using Morphological Image-Processing Techniques for Flexible Pavements

MorphLink-C is a novel image-processing algorithm to connect crack fragments that are a common problem in crack recognition applications. The algorithm consists of two subprocesses: (a) the grouping of fragments by using a morphological dilation transform and (b) the connection of fragments by using a morphological thinning transform. MorphLink-C can be used with various crack extraction methods to connect crack fragments in crack line paths and for complicated crack shapes, such as single cracks, branched cracks, block cracks, and alligator cracks. MorphLink-C also provides a simple but accurate way to estimate an averaged crack width that is important in measuring cracking severity. The proposed method was validated by using realistic road surface images in different pavement cracking conditions. The results of the statistical hypothesis test showed that the proposed method could improve crack detection accuracy with the proposed crack defragmentation algorithm.

Monitoring Proximity Tunneling Effects Using Blind Source Separation Technique

The recent advances in sensing methods and data acquisition technologies have facilitated the collection of instrumentation data for continuous structural health monitoring (SHM). However, interpretation of raw sensor data, affected by various known and unknown environmental factors in field conditions is a challenging task. Structural systems are usually undetermined due to limited sensor data that are not sufficient for finding explicit relations between system inputs and outputs. This study aims to introduce a data-driven methodology using response-only data for underdetermined structural systems. The Principle Component Analysis (PCA) as a Blind Source Separation (BSS) method has been used to decompose the mixed raw response data into a linear combination of statistically uncorrelated mode shapes of input data. Being a data-driven method, the proposed framework is not limited in application to a specific sensor type. To evaluate the efficiency of the method in practice, the close proximity excavation effects of a new tunnel on an existing tunnel has been considered. The analysis results show that the method is not only able to decompose measurements into excavation-induced and the environment-induced deformations but also the calculated eigen-parameters can be used as excellent indicators of structural behaviors during excavation by visualizing the tunnel lining deformations.

Automatic Pavement Object Detection Using Superpixel Segmentation Combined With Conditional Random Field

Pavement images contain various objects, such as lane-marker, manhole covers, patches, potholes, and curbing. Accurate and robust computer vision algorithms are necessary to detect these various objects that have random shapes, colors, and sizes. In this paper, we have addressed the problem of automatic object detection in pavement images using a unified framework. To detect an object of arbitrary shape in an efficient way, we first divide the image into small consistent regions called superpixels. These superpixels are fast to calculate and preserve object boundaries. We then compute several texture and intensity features within each superpixel. After that, we train support vector machine (SVM) classifier for every feature separately in one-verses-all paradigm. In testing, we first estimate the probability of each superpixel being the part of some object of interest using these SVM classifiers. Since these superpixels’ probabilistic scores are independently computed, they do not preserve neighborhood consistency. Therefore, to enforce superpixel neighborhood label consistency, we use contextual optimization technique i.e., conditional random field (CRF). The output of CRF is a pixel-wise binary label map for the objects and background. In addition, due to the lack of any publically available dataset for pavement objects’ detection evaluation, we have introduced a new challenging object detection dataset for pavement images. We have performed extensive experiments on this dataset and have obtained encouraging results.

Statistical Selection and Interpretation of Imagery Features for Computer Vision-Based Pavement Crack–Detection Systems

AbstractThis paper aims to explore the statistics of pavement cracks using computer-vision techniques. The knowledge discovered by mining the crack data can be used to avoid subjective crack featur...