Working with Point Clouds - Cleaning your point cloud

Contents

• Step 1: Copy your point cloud file
• Step 2: Open in Emesent Aura
• Step 3: Make your point cloud visible
• Step 4: Start with a small area
• Step 5: Use the SOR filter
• Step 6: Do a manual clean-up
• Step 7: Export

Every user will clean their point clouds slightly differently. We recommend the following process.

Step 1: Copy your point cloud file

Start by making a copy of your original point cloud file. This will be the file you work on in Emesent Aura.

Step 2: Open in Emesent Aura

Open the copy file in Emesent Aura. You can do this in one of three ways:

• In the top-left menu, click the Project Menu icon then select Open in the menu that displays.
• Drag and drop your file directly into the Viewport.
• Go to the Visualize tab then click Add next to your chosen section. 

Step 3: Make your point cloud visible

To make sure that you can easily see all points for cleaning purposes, we recommend that you do the following.

1. Change the point cloud to a solid color:

   1. Select the point cloud to display its Context panel.

   2. In the Color Scale field, click Solid.

   3. In the Fill Color section, choose a color for your point cloud that contrasts with sepia (which is the default selection color).
      Note: When you change the color of the point cloud, the color of the bounding box automatically changes at the same time. This is useful if you have multiple point clouds open. It allows you to see the extent of each point cloud (assuming they are different colors).

2. Change the background color to a solid color:

   1. In the top-left corner, click Preferences.

   2. In the Preferences dialog box, go to the Appearance tab.

   3. In the Background section, select Solid.

   4. Choose a color that stands out against the point cloud color. We recommend that you avoid using a black background for clean-up, as it can make points easier to miss (especially if you are working with a colorized point cloud)

   5. Click Save.

3. Change your point size so that the points are easily seen:

   1. Go to the point cloud panel.

   2. In the Point Size field, choose a size of 1 or greater. If you use pixel point size, you won’t be able to see your selection.

Step 4: Start with a small area

We recommend that you clean only small areas of your point cloud at a time (particularly if you are cleaning using filters). If you clean the whole point cloud at once, you may accidentally remove features that you didn’t intend to, and there is currently no undo function for deletion.

1. Select a small area of your point cloud. If this area is on the edges of your point cloud, make sure that it includes at least a portion of the main point cloud, not just the peripheral points. The reason for this is so that Aura has a better idea of the mean distance between all the points, not just the peripheral ones. If you only select the points on the edges, this could skew your filtering.
   Note: When you select an area, you will get an extruded selection that will go all the way back through the point cloud. Your selected area includes points behind your visible selection, and these might be areas that you don’t intend to select. If you have chosen perspective mode in your global preferences, the shape of this selection will change (expand) as it moves further into the distance. If you have chosen orthographic mode, the shape will remain the same all the way back through the point cloud.

2. To find points that have may have been selected further into the distance, you can either:

   • Navigate to the other side of your point cloud to check that nothing else has been selected.

   • Use the near clipping plane, which will allow you to see through areas of your point cloud. To do this, go to the Preferences menu and change the value in the Near Clipping Plane field. Experiment with what works best for you.

3. Refine your selection as follows:

   • To add points: Shift + select

   • To remove points: Ctrl + select

   • To invert your selection: Alt + select

Step 5: Use the SOR filter

We recommend that you use the SOR filter for your first clean, before you do your manual point removal. This is a good way to clean up the edges of your scan, where there is usually quite a bit of noise.

The SOR filter allows you to remove stray points and noise in a dense point cloud. This filter works out the average distance of each point from its neighboring points. It then rejects the points that are farther than the average distance. All points outside of this distance are considered outliers and can be removed from the dataset. You may have to experiment with the settings to achieve the desired result.

1. Once you have selected your area for cleaning, go to the Main Toolbar.

   Note: If no selection is made, the filter is applied on the entire point cloud that is currently selected. The filter is disabled if multiple point clouds are selected in the Visualize tab.  
2. Click the Cleaning Filters icon then select SOR filter.

3. In the Statistical Outlier Removal dialog box, select the SOR filter to use. Go to the Main Toolbar section of the Emesent Aura user manual for more information on the different SOR filters and their associated settings. 

4. Once you have configured the settings for the selected SOR filter, click OK.

5. Points selected for clean-up will change to a sepia color. If you are happy with the selection, you can go ahead and delete these points.

6. Run the SOR filter in the same area until you are happy with the result.

Once you have finished using the filters for your initial clean-up, you can then do a manual clean-up. Go to the Main Toolbar section of the Emesent Aura user manual for more information about each tool.

 Step 7: Save

From the Project Menu, click Save to save the changes to the existing file. Use Save As to create a copy of your point cloud with a different name, location, or file format.