Accurate Vergence Control in Complex Scenes

In binocular visual systems, vergence is the process of directing the gaze so that the optical axes intersect at a sur-face point. Correlation-based methods of disparity analysis provide fast estimates of the vergence errol: Unfortunately, most correlation techniques do not provide mechanisms to determine which image locations contributed to Q given cor-relation peak. The result is that large correlation peaks may have contributions from image areas not relevant to the ver-gence task. This paper presents a vergence system that applies a cepstral filter to multiscale images obtained from a dominant-eye binocular sensol: As used by this system, the cepstral filter has two main advantages: it enhances targets through narrow-band signal suppression, and it supports a back-projection operation to determine the image locations associated with particular correlation peaks. The use of multiscale images allows the system to have both high reso-lution for precision in the final vergence and a large field of view for a wide range of initial camera orientations without undue computational cost. 1