| dc.description.abstract | The use of drones has already been utilized in forest inventories with benefits, low cost and high flexibility. This is especially useful considering the difficulties of inventorying in mangrove forest areas because of the vast areas of high-density mangroves. This study aims to mapping mangrove forests using UAV technology and to identify mangrove species from aerial photo in Tanjung Rejo Village, Percut Sei Tuan District, Deli Serdang Regency, North Sumatra Province. The methods used were aerial photography and object-based image analysis (OBIA). The processing of aerial photography into Orthomosaic through the steps includes import photos, align photos, build dense cloud, build mesh, build texture, build digital elevation model, build orthomosaic and export orthomosaic in TiFF format. Object-based image analysis (OBIA) involves segmentation, classification and accuracy testing. The results of the research showed that there are 836 aerial photos at a height of 80 m with the total accuracy of aerial photo mapping resulted in X error of 9.78742 cm, Y error of 7.37204 cm and Z error of 0.0290686 cm. There are 6 types of mangroves identified from the UAV aerial photo with the OBIA method namely Avicennia alba, Avicennia marina, Rhizophora apiculata, Bruguiera hainesii, Nypa fruticans and Bruguiera parviflora. On a segmentation scale of 50 results in overall accuracy value at 78.30%. And on a segmentation scale of 100 results in overall accuracy value at 76.89%. | en_US |
