A system has been developed to enable a robot vehicle to autonomously explore and map an indoor environment using only visual sensors. The vehicle is equipped with a single camera, whose output is wirelessly transmitted to an off-board standard PC for processing. Visual features within the camera imagery are extracted and tracked, and their 3D positions are calculated using a Structure from Motion algorithm. As the vehicle travels, obstacles in its surroundings are identified and a map of the explored region is generated. This paper discusses suitable criteria for assessing the performance of the system by computer-based simulation and practical experiments with a real vehicle. Performance measures identified include the positional accuracy of the 3D map and the vehicle's location, the efficiency and completeness of the exploration and the system reliability. Selected results are presented and the effect of key system parameters and algorithms on performance is assessed. This work was funded by the Systems Engineering for Autonomous Systems (SEAS) Defence Technology Centre established by the UK Ministry of Defence.
This paper addresses the task of making a mosaic from images gathered by a down-looking camera on an airborne platform. This is in the context of a system to detect and map the positions of moving objects. We present three mosaicing approaches based on integrating together sets of measured pairwise homographies, i.e. geometric relationships, between overlapping image frames. The methods are simple chaining, consensus placement and bundle adjustment. We have demonstrated all the approaches with simulated data whilst the simplest way of using pairwise links, simple one-dimensional chaining, has been demonstrated with real data. In our bundle adjustment method, we use a two-dimensional network of pairwise links; when each frame is added to the mosaic, all the constituent frames are adjusted with respect to each other so that the consistency over the entire network is optimised. We have successfully shown, in simulation, that the bundle adjustment technique results in much more consistent, undistorted maps.