Center for Cognitive Sciences

Institute for Neurophysics

University of Bremen


The different perspectives of our two eyes lead to slight relative displacements of objects (
disparities) in the two monocular views of scene.

The human visual system is able to

The correct and fast estimation of disparities is a difficult problem. Besides disparities, various additional image variations occur between the left and right view of a scene. Differences might be caused by occlusions of objects, specular reflections, which move independently of the surfaces of objects, sensor noise, and various other causes.

Stereovision has long been researched; common computational approaches include feature-based, area-based and phase-based methods. All these methods have their intrinsic problems, caused by the very assumptions inherent in these approaches.

On these pages, a simple and fast algorithm is described, which is very different from these classical methods. It is based on the detection of coherence between simple disparity units modeling complex cells in human visual cortex. The neural network implementing the algorithm has some unique properties: it has an internal verification measure, indicating for example areas of occlusion, and it creates as a by-product of the disparity calculations the cyclopean view of the scene.

Coherence-Detection and Disparity Maps by Spiking Neurons

More infos:

- Please note that patents are pending for some of the algorithms described here. -

© 1994-2004 - all rights reserved.