PAVEMENT SURFACE DISTRESS EVALUATION USING VIDEO IMAGE ANALYSIS                 Primary Researchers: Jesus-Adolfo Acosta Robert L. Mullen Department of Civil Engineering Case Western Reserve University, Cleveland, OH 44106, USA E-mail: rlm@po.cwru.edu, Tel. (216) 368-2423 J. Ludwig Figueroa Department of Civil Engineering Case Western Reserve University, Cleveland, OH 44106, USA E-mail: jlf@po.cwru.edu, Tel. (216) 368-6247   Abstract:                                        Maintenance and repair of the highway network accounts for one of the major expenses in the federal and state budget. Pavement Management Systems (PMS) have been implemented by Departments of Transportation and other transportation agencies to optimize the allocation of these funds. One of the most important inputs to a PMS is the pavement surface evaluation. Rating systems where pavement distress is measured by type, extent and severity, have been used extensively in order to quantify pavement surface condition. In most instances, these systems are both tedious and time consuming. Distress measurement is also subjective, which affects the precision of the rating. Identification and quantification of distress types are possible by automatic analysis of images captured by a microcomputer from video or film recordings. The present research describes the implementation of the PCR-Video System, which allows theidentification and classification of most common pavement distress types. Depth and distancemeasurement devices were installed in a survey vehicle and connected to an on board microcomputer to determine the distance traveled and to allow the identification and quantification of depth related distress types. A bar color code method was developed to inscribe distance and depth readings onto the video tape. A functional video image analysis equipment composed of a computer controlled S-VHS tape player, an image capturing board and a workstation was assembled. A set of images is digitized by the image capturing board and stored in main memory to remove overlapping areas present in consecutive frames. The Vertical & Horizontal Region Segmentation method was developed to eliminate the drawbacks found in conventional image segmentation approaches. A logic- based classification approach was also developed for cluster classification. The system when combined with a rating procedure, such as the PCR produces a quantitative measurement of pavement condition. Finally, the pavement inventory data file can be updated with the new pavement ratings. The system was validated by rating four roadway sections, previously inspected manually. The automated results showed very good correlation with the visually obtained ratings.       Typical pavement image with cracking Image of pavement with defects highlighted                                                                          Truck used for testing  CWRU Department of Civil Engineering Main Menu | Sitemap | Department Activities | On-Campus Menu   Communication Click to request a Graduate | Undergraduate Application Form. Click to get in touch with a Graduate | Undergraduate Student. Click to Schedule a Tour of the Department.