Capturing Village-Level Heritages with a Hand-Held Camera-Laser Fusion Sensor

Abstract

In this paper, we present a hand-held fusion sensor system including calibration, motion estimation, and accumulated error reduction for 3D reconstruction. The proposed system consists of four cameras and two 2D laser scanners to obtain a wide field-of-view. The calibration method successfully achieves a much lower reprojection error compared to previous one. The motion estimation method provides very accurate and robust relative poses by fully utilizing plenty observations. At the last stage, the error reduction removes the drift occured over tens of thousands frames with weak GPS prior. Therefore the system is able to capture and geo-register very large heritage architectures spread over square kilometers area. Furthermore, because no assumption or restriction is made, the user can freely move the system and can control the level of detail of the e-heritage without any effort. To demonstrate the performance, we captured 'Hahoe Village' which is one of the most important folk materials of Korea. The experimental result shows that the estimated route fits Google's satellite image while the detailed appearances of representative constructions are captured and preserved well.