A data fusion method for improved motion estimation

Previous work has shown that information from different artificial vision approaches to the same problem can be combined using Data Fusion methods to produce more robust results. Different techniques may also have different computational requirements, and there is often a trade off between algorithm complexity and accuracy. This paper reports the use of state and error predictors combined with data fusion methods to create vision systems with improved performance. Results for both synthetic and real examples are reported. These results show that substantial improvements can be made using this temporal-data fusion approach.