Extracting and Converting a Portion of a Large Point Cloud into a Polyhedral Mesh

From the Civil Engineering section of the action bar, click Extract Polyhedral Large Cloud of Points

.

In the Selections area, you can select different zones of the point cloud to be extracted.

Select the representation reference for point cloud in the tree.

If a cloud of points exists in the 3D Part, it is selected by default.

The Point Cloud Context box is filled in with the representation reference of the point cloud, and the Target Context box with the 3D Shape name, which enables additional selections.

When a hidden area has been defined with the Hide Large Point Cloud command, the Point Cloud Context box is filled in with the name of the Hide feature, Hide.1 for example, and its representation.

To select a sub-part of the point cloud, click New Selection.

The 3D bounding box is displayed and a feature called Bounding Box Trap, Bounding Box Volume or Bounding Box Alignment appears in the tree.

Choose a selection mode:

To make a trap selection in the 3D area, click Trap Mode Selection .
To select a type of trap, click Volume Selection . You can resize the trap in the 3D area.
To select a trap along an horizontal alignment, click Alignment Selection .

If the large point cloud has LIDAR classification, the extract will use a set of classification different from the one used in the visualization.

Notes:

You cannot create a contextual feature for the bounding box. For the volume and alignment bounding box, existing geometry is required.

To select the classification that the extract will take into account, click

.

If there are at least two classifications that are valued in the large point cloud, this button is available. When clicking on it, a panel with a classification list is displayed. Only the classifications containing points are displayed.

To manage the number of selected points, you can use a filter in the Options area of the Extract Large Point Cloud dialog box, and select

The Homogeneous method: To reduce the number of selected points, you can filter the points with a minimum distance between points, for example: 10mm means that the points, at a distance below 10mm, will not be selected.
The Adaptive method: The extract output has more points around the high gradient area and less points in the planar area.
This mode is based on the local chordal deviation. It keeps points around areas which are not plane and discards points in planar areas. You cannot have the two types of filter activated at the same time. It is either an adaptive or an homogenous filter.
The performance on this filter depends on the number of points of the large point cloud within your selection before filtering.
When you set the value to zero, the filter is disabled. The Maximum Distance parameter is available, only for the Adaptive mode, to increase the number of points kept in planar area. It specifies the maximum distance between two points to be considered during extraction, up to which gaps are not visible. A certain density of points is required for the triangle creation. When the Maximum Distance is selected, the mesh triangles are smaller.

The option is useful for alignment selection.

When selecting a mode, the box next to it can be edited and the other one is disabled.

If you edit the same feature, the value of the disabled filter will be zero.

To select the percentage of the outlier points to be ignored, drag the slider in the Options area of the Extract Large Point Cloud dialog box.

The percentage slider corresponds to the number in percentage of the total outlier points that can be ignored in the mesh. You can only use these predefined values: 0%, 25%, 50%, 75%, 100%. If 100% is selected, all the outliers points are discarded by default.

Since the data should represent terrain, the direction used to calculate the distance is the Z coordinates.

Note: The 3D Mesher computation discards some points that it judges to be outliers. This behavior is based on the surroundings points of the current point. If the point is far away from the surrounding points, it is discarded. Hence, it is not necessary to remove all the outliers.

Optional: Select the Compute the number of points that will be contained in your selection. Each time you change a parameter of the extract, the point cloud is computed again. It may be time consuming depending on the size of the large point cloud.

To specify if the extracted point cloud should contain the points inside the alignment volume and outside the trap for example, select Inside or Outside in the Trap Option column of the Extract Large Point of Cloud dialog box.

If satisfied with your selection, click OK.

Optional: To edit a selection, double-click the bounding box feature in the tree or click a row in the list and click Edit Selection.

You get the box in the 3D area and you can change the dimensions.

Optional: To edit the selection, double-click the bounding box feature in the tree or click a row in the list and change the dimensions.

To delete a selection, click Remove Selection.

Optional: Click Apply to preview the result.

To run the Extract command, click OK.

The cloud of points corresponding to the points defined by the area is extracted. You obtain a polyhedral mesh in the 3D area. In the tree, the Extract feature is created under the Geometrical Set node with the polyhedral mask . It contains a sub-item for:

The Homogeneous or Adaptive filter mode.
The Maximum distance, if defined.
The Bounding Box type.

Associativity is kept between the cloud of points and the mesh.

Notes:

The mesh does not have the color of the extracted points. The default color of the polyhedral mesh is used, but you can edit.
A progress bar informs you about the progress of the computation. Press Escape to cancel if needed.

To synchronize the extract feature with the cloud of points, click the feature in the tree and select the Extract object > Synchronize contextual command.