Machine vision using a combination of optical sensors with different measurement accuracies and fields of view, so called multi-focal vision, is investigated on different abstraction levels: static, dynamic, and planning. The performance of multi-focal vision systems is quantified. Novel multi-focal vision-based and attention control strategies are proposed, analyzed, and compared to conventional strategies. These novel approaches enhance the perceptual and controlling capabilities of Robots by combination of large field of view, improved measurement accuracy and control performance as well as a flexible situative allocation of sensor resources.
«
Machine vision using a combination of optical sensors with different measurement accuracies and fields of view, so called multi-focal vision, is investigated on different abstraction levels: static, dynamic, and planning. The performance of multi-focal vision systems is quantified. Novel multi-focal vision-based and attention control strategies are proposed, analyzed, and compared to conventional strategies. These novel approaches enhance the perceptual and controlling capabilities of Robots by...
»