This paper explores integrating Geographic Information System (GIS), Computer Vision (CV) and Artificial Intelligent (AI) using cellular phones to evaluate and manage pavement conditions. It enables non-expert users to assess and maintain pavements effectively. The study develops CV-based mapping and a knowledge-based system for distress detection and management. Statistical regression models predict pavement sustainability. Finally, this approach empowers users to make informed maintenance decisions. Different cameras resolutions ranging from 0.2 to 16 MP were used to build an intelligent framework system to evaluate and manage pavement conditions with suitable maintenance works by non-expert users. A CV based system was developed for pavement surface mapping using cellular phones. A macro scale mapping of pavement surface conditions in the presence of flexible pavement distresses was developed by cellular phones with normal based configuration and various camera resolutions of arterial roads in Irbid city, Jordan. GIS layers were built for pavement conditions with various parameters, rating, distress types, severities and repairs options based on Global Position System (GPS) determination for distresses locations. The developed GIS System was established by integrating a set of computerized programs as a part of GIS software. New parameters were introduced to the system to expedite the pavement distresses classification, detection, management, and maintenance process, taking into account distress types, severities and geometrical measurements. A knowledge-based system (KBS) for pavement maintenance was also developed. It took into consideration distress type, severity, and pavement conditions. A criterion for images enhancement processes based on image processing technique for pavement distresses detection and management was developed. Surface measurements, pavement conditions as well as decision-making tasks have been supported and considered for all distress types. The developed statistical regression analysis models for pavement sustainability, serviceability and condition prediction utilized a set of extracted resulted measured variables of Pavement Condition Index (PCI), Present Serviceability Index (PSI) and Sustainability Index (SI) from various sources. They were conducted for the collected actual data with the values of (84.8, 2.988 and 0.6057) respectively. Regression results were statistically significant for all models with normal probability plots near a straight line.
T. Obaidat, Mohammed; W. Al-Mestarehi, Bara; and H. Bani Ata, Tamara
"Digital Mapping of Urban Arterial Roads Pavement Conditions,"
Information Sciences Letters: Vol. 13
, PP -.
Available at: https://digitalcommons.aaru.edu.jo/isl/vol13/iss1/13