Development of Photogrammetry Application for 3D Surface Reconstruction

Abstract

Photogrammetry has long been used to collect three-dimensional (3D) information about objects or texture data. The 3D reproduction of real-life objects can be highly useful in a variety of fields, such as archeology, medicine, and geology. The advantages of photogrammetry research with lasers reduce costs, time to digitize objects, and have high accuracy and efficiency. This research aims to design an application that processes 2D image input data into 3D. The features contained in this application include data collection, calibration, image rectification, feature detection, and scanned images. The development of this application has succeeded in reconstructing 2D images into 3D images. In the test, the average visible point detection point is 185.4 and the mean re-projection error is 0.7542. The minimum calibration error must be obtained so that the surface reconstruction results are close to the actual shape. In the future, we would extend our work by manipulating robot movement with photogrammetry as input data.