Vergence error extraction and vergence servo control is important issu
es in gaze stabilization. In the previous paper, ''Purposeful Gazing f
or Active Vision I: Vergence Error Extraction'', we have discussed the
techniques in fixation point selection and the phase-based vergence d
isparity extraction algorithm. In this paper, we introduce a feedback
information prediction and dynamic vision-based self-tuning control st
rategy to achieve vergence control. This strategy reflects the nature
and function of visual processing involved in vergence control loop an
d tends to correct some confusion in existing vision-based control sys
tem development. (C) 1997 Pattern Recognition Society. Published by El
sevier Science Ltd.