Pleiades satellite image object-based classification and accuracy analysis with airborne lidar data integration

Remote sensing technologies have many methods for land cover mapping. With the increase in urban development, geographical databases need to be used and updated accurately and effectively. Object based classification is an effective classification method for producing land use maps from high resolution satellite images. The high resolution products from airborne lidar technology, which allows the topography to be acquired very quickly and precisely, can be integrated into object based classifications to increase classification accuracy. In this study, high resolution Pléiades satellite image was initially applied to object-based classification using only spectral data and then integrated high resolution DSM, nDSM, intensity dataset which generated from airborne point cloud data to analyze the change in the classification accuracy. As a result of the analysis, it has been found that combining spectral data provided by satellite images with 3D spatial products obtained from airborne data increases the classification accuracy, especially in the urban areas. Classification overall accuracy and Kappa statistics are incrased 6 % and 0,09 respectively.