City information model semantic information artificial intelligence generation method and system
A technology of semantic information and artificial intelligence, applied in the field of 3D modeling, it can solve the problems of cumbersome process, heavy tasks, and inability to define object attributes separately, so as to achieve the effect of convenient acquisition and improved extraction efficiency.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0064] In S1-1, during the 54,000 sunny morning moments, one remote sensing image of city A, one corresponding laser scanning point cloud image, and one corresponding DSM image were simultaneously obtained at each moment, totaling 54,000×3 The image data forms multiple sets of data for modeling. Among them, the data used for modeling between 10 groups come from different geographical ranges of city A (that is, the data obtained come from different areas of city A).
[0065] Such as figure 1 , where there is a complete remote sensing image of a municipal building with a rectangular roof frame, a corresponding corresponding laser scanning point cloud image, and a corresponding corresponding DSM image;
[0066] In 1 complete remote sensing image of a municipal building with a rectangular roof border, the corresponding 1 corresponding laser scanning point cloud image, and the corresponding 1 corresponding DSM image, a roof vertex of a municipal building with a rectangular roof bo...
Embodiment 2
[0073] S2-1 Obtain a predetermined number of roof interiors in each of the remote sensing image layers of the multi-layer 3D model in the training set and are close to (for example, the distance from a point on the contour is 1 / 3 of the distance between the contour point and the geometric center of the contour) The RGB three-color value of the pixel at the contour of the inside), and each remote sensing image layer is divided into multiple blocks. Such as Figure 5 shown will be figure 2 The medium and remote sensing image layer is divided into 16 blocks, and there is no roughly parallel building outline near the boundary of each block as much as possible. It should be noted here that the block is divided into rectangles and does not necessarily need to be divided into equal parts. S2-2 Obtain the clustering diagram of different roof materials (including square ground) RGB three-color value classification (see Table 1 for typical RGB values) through clustering algorithm (su...
Embodiment 3
[0085] S3 specifically includes:
[0086] S3-1 According to the distance information of all the characteristic geometric surfaces A in the remote sensing image layer in the multi-layer 3D model in the test set, translate all the characteristic geometric surfaces A along the Z-axis direction of the coordinate system established in the layer to The corresponding height H,
[0087] S3-2 establishes a database on all the characteristic geometric surfaces A after translation, so that when the multi-layer three-dimensional model is displayed, the mouse can be moved to any characteristic geometric surface A or any characteristic geometric surface A can be displayed. The identification information, geometric information, and material information of a characteristic geometric surface A, together with the artificial intelligence model M and the roof material classification model S f , multi-layer 3D model to jointly generate the city information model CIM (the process is as follows F...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More