Information-based visual robotic mapping in unstructured environments

In field environments it is not usually possible to provide robotic systems with valid geometric models of the task and environment. The robot or robot teams will need to create these models by performing appropriate sensor actions. Here, an algorithm based on iterative sensor planning and sensor redundancy is proposed to enable them to efficiently build 3D models of the environment and task. The method assumes stationary robotic vehicles with cameras carried by articulated mounts. The algorithm uses the measured scene information to find new camera mount poses based on information content. This method is particularly well suited to unstructured environments, where sensor uncertainty is significant. Issues addressed include model-based multiple sensor data fusion, and uncertainty and vehicle suspension motion compensation. Simulations and experiments show the effectiveness of this algorithm.