A method of object detection

Abstract

The invention relates to a method of object detection in which two images of an environment (1) and a disparity image is determined by stereo image processing, wherein the disparities determined a depth map of the area (1) is determined from where a free space boundary line (2) identifies is that defines an unobstructed area of ​​the area (1), wherein the depth map is segmented outside and along the clearance gauge (1) by segments (3) are formed of a suitable width of pixels of the same or similar distance to an image plane, wherein a height each segment (3) as part of an object located outside the obstacle-free area is estimated (4.1 to 4.6), so that each segment (3) is (given by range and angle to the longitudinal axis, for example) and its height characterized by the two-dimensional position of his foot point ,