Control issues in active vision: open problems and some answers

Abstract

Hardware developments have rendered controlled active vision a viable option for a broad range of problems, spanning applications as diverse as intelligent vehicle highway systems, robotic-assisted surgery, 3D reconstruction, inspection, vision assisted grasping, MEMS microassembly and automated spacecraft docking. However, realizing this potential requires having a framework for synthesizing robust active vision systems, capable of moving beyond carefully controlled environments. In addition, in order to fully exploit the capabilities of newly available hardware, the control and computer vision aspects of the problem must be addressed jointly. We illustrate with a simple example the control-related issues involved in active vision and we show how some very recently developed control and computer vision techniques can be brought to bear on the problem. These results also point out new research directions and possible extensions of currently available techniques.