| 摘要 |
The invention relates to a method of acquiring simulated autostereoscopic video images of a scene to be viewed. On the basis of stored data containing three-dimensional information, it implements n simulated cameras, with n>=3, each generating an image of a scene on a given optical axis. The optical axes of the simulated cameras converge at a point situated at the same distance D from the simulated cameras. The scene to be viewed has a nearest point Pp and a farthest point Pe, and the inter-camera distance and the distance Dmin between the simulated cameras and the nearest point Pp are selected in such a manner that for focus varying between the nearest point Pp and the farthest point Pe the angle 2alpha between two adjacent simulated cameras varies between a value not greater than 4.5° for the point Pp and a value not less than 0.2° for the point Pe. |
