A method of detecting an object on the road side of a car, a computing device, a driver assistance system and an automobile

摘要

The present invention relates to at least two images (13) of the periphery (9) of a car (1) continuously captured by at least one vehicle-side camera (4) for external correction of at least one camera The image 13 representing at least in part the texture of the road surface 11 and comprising the steps of: a) detecting the object 12 on the road 10 side of the vehicle 1, Determining at least two mutually corresponding blocks (14) based on at least two images (13); b) determining a respective motion vector for each of at least two pairs of corresponding blocks (14); c) determining depth information associated with at least two images (13) based on at least two motion vectors; d) a step of detecting the object 12 is performed based on the depth information. The present invention also relates to the computing device 3, the driver assistance system 2, and the automobile 1.