2013 Session: 475

• Assessment of Pavement Cracking in Real Time Using Three-Dimensional Laser Technology and Advanced Image Analysis Algorithms
  Abstract: The analysis of pavement distress through cracking has provided major challenges for road authorities. Manual methods of analysis lack consistency due to the subjective nature of assessment and issues with image quality. Recent advances in image analysis and illumination have enabled automation of the data collection process, but have limitations in terms of data processing and classification. This paper compares two automated systems, RoadCrack and the Laser Crack Measurement System (LCMS), and recent advances in the integration of these technologies to provide real time cracking analysis.
  Authors: Warren, Garry
  Authors: Warren, Garry
  Year: 2013
  Document Type: Paper
  Subject: Design; Pavements
  Session: 475
  Paper Number: 13-2434
• Crack Detection in Pavement Images Using Texture Analysis and Unsupervised Learning
  Abstract: The second phase of a two part study on using an unsupervised learning technique for the detection of road cracks from pavement images is described in this paper. The main concentration is on highly textured road images that make the crack detection very difficult. Road images are split into smaller rectangular cells and for each cells a representative data set is generated by analyzing image texture and colour properties. Texture and colour properties are combined with an unsupervised learning technique called the Kohonen map to distinguish crack areas from the background.Using this technique, cracks are detected to an accuracy of 75% was achieved. The background was segmented correctly with more than 99% accuracy despite it having very strong visual texture. The technique applied here shows a great deal of promise despite that the images were captured in an uncontrolled environment devoid of state-of-the-art image acquisition setups.
  Authors: Mathavan, Senthan; Rahman, Mujib; Kamal, Khurram
  Authors: Mathavan, Senthan; Rahman, Mujib; Kamal, Khurram
  Year: 2013
  Document Type: Paper
  Subject: Design; Pavements
  Session: 475
  Paper Number: 13-3174
• Matched Filtering Algorithm for Pavement Cracking Detection
  Abstract: There have been rapid developments of automated cracking survey for pavements in recent years. Especially the introduction of laser-imaging technology has made the acquisition of shadow-free images feasible. However, due to the complexity of pavement surface, diverse characteristics of cracks, presence of foreign objects, and varying identification protocols, the results of fully automated technologies for cracking survey have only been used on limited basis. This paper introduces a novel detection theory, matched filtering algorithm, into pavement cracking detection as part of an effort to improve precision and bias levels of the fully automated system. Unlike traditional edge detection approaches which adopt first- or second-order derivatives of image signals, the matched filtering algorithm detects cracks by matching pre-designed filters with crack features in terms of shape, orientation and intensity. Experimentations are conducted and detection results are compared between the five traditional edge detectors (Robert, Prewitt, Sobel, LoG and Canny) and the matched filtering algorithm. It is demonstrated that the matched filtering algorithm is a robust approach to detecting cracks and has better performance in noise removal and cracking detection. With matched filters aligned at different orientations, this algorithm shows its distinctive advantages at extracting a single crack as it is, and recording the crack’s orientation to be used for more accurate classification in the next step of automated processing.
  Authors: Zhang, Allen; Li, Qiang; Wang, Kelvin C. P.; Qiu, Shi
  Authors: Zhang, Allen; Li, Qiang; Wang, Kelvin C. P.; Qiu, Shi
  Year: 2013
  Document Type: Paper
  Subject: Design; Pavements
  Session: 475
  Paper Number: 13-4077
• Enhanced Crack Segmentation Algorithm Using Three-Dimensional Pavement Data
  Abstract: Automatic pavement crack segmentation has gained attention among researchers and transportation agencies over the past two decades. However, most existing algorithms using 2-D pavement intensity images cannot provide satisfactory performance. With the advance of sensing technology, 3-D line laser pavement imaging systems have become available. The objective of this paper is to propose an enhanced dynamic optimization based algorithm utilizing the advantages of 3-D pavement data to improve crack segmentation. The proposed algorithm consists of three major stages. First, a two-step Gaussian filter is applied to remove outliers from the collected laser data and rectify the profile in order to reduce the influence of cross-slope and ruts on crack segmentation. Then, a rough crack segmentation stage is conducted to adaptively identify the crack regions of interest. Finally, a bounding box and major orientation for each valid crack region of interest will provide searching space and direction for the precise crack segmentation using the dynamic optimization based algorithm. Experimental tests are conducted using actual pavement data collected near Savannah, Georgia. The four most common types of pavement cracking (longitudinal, transverse, block, and alligator cracking) are tested, and the performance between original dynamic optimization algorithm and the proposed algorithm is compared. Experimental results show that the proposed algorithm only take about 1/4 of the average computation time of the original algorithm. Also, the accuracy of crack segmentation has been improved, since the proposed algorithm removes the unnecessary false positives and better handles cracks in multiple directions. Finally, conclusions are made and recommendations for future research are discussed.
  Authors: Jiang, Chenglong; Tsai, Yichang (James)
  Authors: Jiang, Chenglong; Tsai, Yichang (James)
  Year: 2013
  Document Type: Paper
  Subject: Design; Pavements
  Session: 475
  Paper Number: 13-4021
• Assessment of Pavement Cracking in Real Time Using Three-Dimensional Laser Technology and Advanced Image Analysis Algorithms
  Authors: Warren, Garry
  Authors: Warren, Garry
  Year: 2013
  Document Type: Presentation
  Subject: Design; Pavements
  Session: 475
  Paper Number: 13-2434
• Matched Filtering Algorithm for Pavement Cracking Detection
  Authors: Li, Qiang
  Authors: Li, Qiang
  Year: 2013
  Document Type: Presentation
  Subject: Design; Pavements
  Session: 475
  Paper Number: 13-4077
• Crack Detection in Pavement Images Using Texture Analysis and Unsupervised Learning
  Authors: Mathavan, Senthan
  Authors: Mathavan, Senthan
  Year: 2013
  Document Type: Presentation
  Subject: Design; Pavements
  Session: 475
  Paper Number: 13-3174
• Enhanced Crack Segmentation Algorithm Using Three-Dimensional Pavement Data
  Authors: Jiang, Chenglong
  Authors: Jiang, Chenglong
  Year: 2013
  Document Type: Presentation
  Subject: Design; Pavements
  Session: 475
  Paper Number: 13-4021