AU - Mohammadi, Reza AU - Farnaghi, Mehdi TI - Designing and developing a multi-Agent system for automatic extraction of road geometry by the Crawler-Agent PT - JOURNAL ARTICLE TA - kntu-jgit JN - kntu-jgit VO - 5 VI - 2 IP - 2 4099 - http://jgit.kntu.ac.ir/article-1-465-en.html 4100 - http://jgit.kntu.ac.ir/article-1-465-en.pdf SO - kntu-jgit 2 ABĀ  - Up-to-date digital maps play a significant role as required data in various areas such as navigation, tourism, traffic control, urban and interurban fleet management, ITS, web-based map services and location-based services. Nowadays, VGI is known as an important resource to generate digital maps. In the last few years, the use of these resources to produce low cost, up-to-date, and reliable maps have been considered. Previous studies have utilized static methods for automatic road extraction from volunteered trajectory data. However, they have not considered the continuous changes of the road network which are reported dynamically by the volunteered trajectory data of users. So, due to their static nature they are not efficient to update road maps dynamically. Considering the fact that crawler-agents are capable of sensing and reacting changes in the environment it is possible to dynamically update road maps based on the latest changes presented by the trajectory data. The key idea of this research is to design and develop crawler-agents that dynamically search and update road maps from volunteered trajectory data. In this article, inspired by the continues monitoring of the web by web Crawler-Agents, a dynamic method is presented to updating road maps with considering the collected participatory data that has been gathered and recorded by the GPS sensors on the mobile device of users. To do so, various group of agents that move on trajectory data to extract road map are designed and developed. EM clustering is used to estimation of the network nodes. Also, a heuristic method is developed to link the nodes which are related to each other. Furthermore, an efficient approach is presented to extraction of road networks e by applying three dimensional trajectory data in the case of road’s challenge structure. CP - IRAN IN - LG - eng PB - kntu-jgit PG - 35 PT - Research YR - 2017