Low-cost 3D surface reconstruction using Stereo camera for small object

Abstract

3D data needs in various fields has increased along with the development of simulation and modeling technology. The current technology makes interaction with 3D data becomes easier and interesting. The existence of virtual reality and 3D printers provide ease to create and display 3D objects. The existence of a 3D scanner to reconstruct a real object into a 3D digital object is one of the practical solutions to facilitate the creation of 3D data. The methods to perform 3D object reconstruction process also began to be found, such as using a single camera, stereo camera, laser beam, structured light, and the combination of these methods with various algorithms. In this study, we create a prototype of low-cost 3D surface reconstruction system with a stereo camera. The Stereo camera using image processing to generate point cloud of the object with displacement and disparity mapping process. The point of cloud from the object then connected to produce 3D surface. Reconstruction is done by applying texture from object photograph to the surface of 3D point cloud that has been reconstructed. The system made by using a low-cost web camera to test the ability to reconstruct 3D objects surface with inexpensive equipment. Testing is done by giving structured light variable using the projector. The results of this study indicate that the calibration process camera position and light conditions will determine the outcome of reconstruction where the dual camera system is very dependent on the calibration process to produce a proper reconstruction.