| 摘要 |
A method for generating a 3D ground model and a 3D ground surface model in a stereo image are provided to generate the 3D ground model and the 3D ground surface model even in stereo images including artificial objects with problems such as a surface of discontinuity in height, an occlusion area, and different front surfaces, thereby reducing cost and time. A method for generating a 3D(Three-Dimensional) ground model and a 3D ground surface model in a stereo image comprises the following steps of: photographing the same region into plural sheets of stereo images at different locations and angles(a); extracting the roof side and front side of artificial objects in each stereo image from each stereo image(b); removing the extracted roof side and front side of the artificial objects from each stereo image(c); repeating image matching in each stereo image(d); generating the 3D ground model in which the artificial objects are removed from the image matching result(e); generating the 3D coordinates of the roof side of the artificial objects from the roof side and front side of the artificial objects extracted from each stereo image(f); and combining the 3D ground model and 3D coordinates of the roof side of the artificial objects to generate the 3D ground surface model(g). |
