Determination of vehicle with remote sensing and geographic information system in the highway with meanshift, multi resolation segmantation and kmeans technique

Urbanization increase rapidly in the world, cities and urban areas connection with each other is so vital important subject. Highways connect residential area with other where. Roads respond to increasing intensive population. For this reasons they are renewed and enlarged. A good transportation in urban area depends on the healthy operation and use of roads. In our country, increase of urbanization, many metropolitan cities are formed and the transportation between these metropolitan cities and settlements is vital. Highways are one of the most important components in urbanization and urban planning. Traffic is a very important criteria for planning and construction. Traffic is defined as the traffic law of highways and all movements of pedestrians, animals and vehicles on highways. Vehicle is most important component to traffic. Traffic has become one of the most important problems especially in many metropolitan and residential areas in recent years. For this reasons highway investment reach very much value. Geographic information systems and remote sensing are used urban and transporting planning. In this study, it is aimed to determine the vehicles on road by means of accessible data sources. Google maps, Yandex maps, Bing maps exc. data source accessible from everyone. These data are used in areas such as navigation, mapping, and updated at regular intervals. The aim of the study is to determine the tools by applying the segmentation procedures based on the principle of grouping of objects with certain features which are remote sensing techniques. The methods of segmentation, mean shift and object-based multi-resolution methods were applied. In addition to these two methods, uncontrolled classification was made and compared with the results of segmentation. The study was completed with the help of some GIS techniques. As a result, it was calculated that road vehicles were detected with 80% accuracy