A Portable Multi-sensor System for Geo-referenced Image Sequences

The demand of high-resolution spatial information such as geo-referenced images and realistic 3D models is rapidly increasing. Most existing systems to acquire such information involve expensive platforms such as manned or unmanned, air or ground vehicles to obtain high-resolution images and GPS/IMU data. Thanks to the availability of small image acquisition devices and light-weighted MEMS GPS/IMU in conjunction with a portable UMPC, it is possible to obtain spatial information without any platform but a person. In this study, we developed a Portable Multi-sensor System consisting of a video camera, a MEMS GPS/IMU and a UMPC to acquire video images and position/attitude data. We performed image geo-referencing based on the bundle block adjustment without control points and evaluated our system by verifying the accuracy of object points derived from the geo-referenced images. The experimental results show that the RMSE of relative coordinates of object point was only a few centimeters. Our system can be efficiently utilized to obtain the 3D models of objects. In the future, we plan to improve the accuracy of absolute coordinates through the rigorous calibration of the system and camera.