DECISIONS IN PREDICTIVE LOCATION AND MAPPING
- Additional Document Info
- View All
This paper presents our recent progress in simultaneous localization and mapping. Discussing recent progress in algorithms for dense reconstruction of unstructured environments and robust self-navigation, we present our work along with a variety of applications. In the light of this discussion, it is felt that the fusion of computational vision with control can now be contemplated on the near horizon. Key challenges that enable this information driven fusion are presented in the paper. Some preliminary analysis and associated results are presented on the topics of bundle adjustment, feature based navigation and object space tracking to report progress in addressing the key challenges identified. Some experimental test results are presented to document the efficacy of the algorithms presented in the paper. Prefatory experiments on kinect based proximity spacecraft motion emulation experiments are shown to demonstrate the challenges outlined in the paper.
author list (cited authors)
Majji, M., & Junkins, J. L.