| 摘要 |
Apparatus and method for image enhancement in which an autostereoscopic image, comprising a different perspective view of an object field for each eye of an observer is, typically provided. The optical cross-talk, which is often a feature of autostereoscopic images, is suppressed when images are displayed, by preconditioning of the image intended for each eye. Supplementary image data is produced by mixing the different perspective image data and this is applied to the primary data. By inversion of a cross-talk matrix, two or more different perspective channels can be preconditioned to cancel the effects of such cross-talk, providing clear images for each eye without the unwanted interference from other views. The viewer's location with respect to the display device can be used to modify the characteristics of both the display and the level of cross-talk correction. A camera or endoscope for providing stereoscopic images can also suffer from optical cross-talk. The imagery captured by such a system may be improved by applying the same principles as are applied to the display process. |
