Planning viewpoints that simultaneously satisfy several feature detectability constraints for robotic vision

Abstract

The authors present results of a method developed which automatically determines viewpoints for a robotic vision system so that object features of interest are simultaneously visible, inside the field-of-view, in-focus and magnified as required. The authors previously (1990, 1991) analytically characterized the domain of admissible camera locations, orientations and optical settings for which each of these feature detectability requirements is satisfied separately. In addition, they presented a technique that poses the problem in an optimization setting in order to determine viewpoints that satisfy all requirements simultaneously and with a margin. In this paper, they present results of this technique when all the feature detectability constraints are included. In order to verify that all the constraints are indeed satisfied, camera views are taken from the computed viewpoints by a robot vision system that is positioned and its lens is set according to the results of this method. >