| 摘要 |
The present invention relates to determination of a terrain parameter relative to a vehicle (100) by registering a first image from the vehicle (100) when it is located at a first point in space (O1). The first image includes a primary representation of a first terrain point (P1) and a primary representation of a second terrain point (P2). Then, a second image is registered from the vehicle (100) when it is located at a second point in space (O2). This second image includes secondary representations of the first and secondary terrain points (P1) and (P2) respectively. Moreover, vehicle movement data is received, which describes at least one of a translation (R) between the first and second points in space (O1; O2), and a rotation (A) between the first and second points in space (O1; O2). A first image velocity vector is determined, which describes a movement of the primary representation of the first terrain point (P1) in the first image to the secondary representation thereof in the second image. Correspondingly, a second image velocity vector is determined, which describes a movement of the primary representation of the second terrain point (P2) in the first image to the secondary representation thereof in the second image. A terrain vector ( XI ) is then resolved, based on the first and second images, a camera model describing a geometric relationship between a particular terrain point and its projected representation onto an image plane, and a respective relationship expressing each of the first and second image velocity vectors as a function of the translation (R) and the rotation (A). The terrain vector ( XI ) extends between one of the first and second points in space (O1; O2) and a particular terrain point (T) in proximity of the vehicle (100). Thus, based on the terrain vector ( XI ), for instance an altitude (ZO1) of the vehicle (100) may be determined. <IMAGE>
|
