An integrated vision-based positioning system for video stabilization and accurate local navigation and terrain mapping

This paper deals with the design, construction and calibration processes in the use of angle sensors to improve the accuracy of a vision-based positioning system. These sensors are used for measurements that cannot be determined accurately (enough) from motion vision techniques. Furthermore, they provide a mechanism to stabilize the video data with respect to undesirable motion effects, e.g., cyclic motions in the surf zone. To use the integrated system effectively, a high-precision calibration procedure is employed to determine the transformations among the coordinate systems of the angle sensors and the vision system. With the positioning information presented in a common coordinate frame, the ROV control system can be optimized for operations that require high-precision navigation, trajectory following, terrain mapping, etc. Results of the calibration process and the testing of the system are presented to evaluate effectiveness and performance.