Komparasi Model 3D Jembatan dari Pemotretan Konvergen dan Normal Menggunakan Drone

Abstract

CRP (Close Range Photogrammetry) is a measurement technique that can extract 3D points from an image. This study aims to compare convergent, normal, and combined shooting using drones with close-range photogrammetry methods. In the CRP technique, shooting is carried out using the DJI Phantom 4 Pro Drone with convergent and normal shooting techniques and a combination of both shots. Convergent shooting is done by positioning the drone centered on one point by forming a cone against the bridge object with an angle ranging from 40-90 degrees. While shooting normally is done by positioning the drone by facing a straight line against the object. A comparison of concurrent and normal shooting was carried out to find out better 3D Modeling results. The final result of this research is the visualization of a 3D model object that has a complete texture and resembles its original state in the field, as evidenced by the RMSE value of the 3D convergent model being 0.0940 m, for the RMSE value of the normal 3D model is 0.0882 m. The RMSE 3D value of the combination model is 0.0907 m. From each RMSE value above, a good 3D model visualization is produced using normal shooting because the resulting RMSE value is smaller than other shooting techniques.

Keywords

• close range photogrammetry
• convergent
• modeling
• normal