Mersin Photogrammetry Journal

Yazarlar: Abdurahman Yasin YİĞİT, Murat UYSAL

Konular:-

Anahtar Kelimeler:Object-based,Automatic Road Detection,Classification,UAV,Road

Özet: Aerial photos and satellite images tell us about the land surface. It provides a variety of information such as particularly human-built objects such as buildings, roads and bridges and the location and characteristics of the vegetation. With the exception of aerial photographs and satellite imagery, the collection, evaluation, and updating of the required data with other data collection methods is a time-consuming process and more costly. Data from aerial photographs and satellite images have long been detected manually by conventional methods and by operators. The automatically of these detections increases the speed of the project process and contributes to the reduction of the expenses spent. The projects carried out within the scope of the extraction and classification of objects are mostly concentrated on buildings and roads. Because roads and buildings; Due to the characteristic features such as having sharp lines and easy determination of the geometric shape, the identification of the detail lines in the objects is easier than determining the details of other objects. Various object extraction and classification techniques are used for image analysis with semi and fully automatic approach methods based on image processing techniques. In this study, aerial photographs of a certain area of Afyon Kocatepe University Ahmet Necdet Sezer Campus were obtained by using unmanned aerial vehicles (UAV). The raw data obtained were evaluated and the object-based classification approach was used to automatic detection and classify the roads of the university in the digital environment.

---