Watch our tutorial on georeferencing Polycam scans using CloudCompare!
Preparing the Point Cloud Capture
We will be working with this captured garden space created by FuturisticIllusion1523. Within the details window, click Download 3D Model, and then choose the Georeferenced LAS file and click Export.
Fixing LAS File Import Issues
Now, if you try and drag and drop a raw LAS file directly from Polycam into CloudCompare, you will run into this issue. The point cloud data appears in a single straight line of points. To fix this, we will use a free program called QGIS to export the data with a specific type of coordinate system that CloudCompare is designed to read. Open QGIS and then drag and drop your LAS file into the workspace. It may take a moment to render the point cloud. That's normal, and we don't need to wait for that all to load. We're only going to use QGIS to convert the file.
Exporting a Converted LAS File
In the Layers panel, right-click your file name and go to Export, Save As, and choose LAS/LAZ Point Cloud Format. Give the file a new name, such as Converted LAS two.
Understanding Coordinate Reference Systems
Next, we need to update the CRS, or coordinate reference system. By default, smartphones and GPS technology uses WGS 84, which is great for determining where you are on the planet. But for CloudCompare, we need WGS 84 UTM Zone CRS, which is more accurate for local distance and area mapping, and it's more typically used in architectural or city planning applications. Click the globe icon to open up the selector and search for three two six three three, which will pull up the WGS 84 UTM zone and select it. Then click OK. This will generate a new LAS file with the correct coordinate system.
Importing the Converted File Into CloudCompare
From there, you can navigate to your file location and find the converted LAS file and drag it into CloudCompare. When prompted in the first window, click Apply All, and in the second window that comes up, make sure that Preserve Global Shift on Save is checked on and click Yes to All. Now you'll see the intended point cloud imported correctly as a new layer in CloudCompare.
Understanding the CloudCompare Interface
Now that the file is imported, let's review CloudCompare's interface. On the top left is the database tree, which shows you your layers. Below it, you'll find view controls for switching between orthographic views like front, back, side, and perspective. A quick reminder here that CloudCompare does not have an undo button. So to protect your work, create a duplicate before making changes. Select your file and then click the clone icon. But for this video, we do not need to clone this mesh, so we'll go ahead and remove it.
Cleaning Noise From the Point Cloud
Now let's get to work on cleaning up the point cloud. With the capture selected, go to Tools, Clean, Noise Filter. We will leave the settings as they are, but we'll make sure that Remove Isolated Points is selected. This filters out the noise and generates a new cleaner layer. You can toggle between the original layer and the cleaned versions in the database tree.
Cropping the Point Cloud
Next, we'll use the segmentation tool to crop the point cloud. Select your layer and click the scissors icon in the top toolbar. This opens a segmentation mode. You can choose either polygon selection or rectangle selection. I'm gonna stay with the more versatile polygon selection tool. Draw a polygon around the area you want to crop. Right-click to close the shape. Then choose whether to crop in or crop out. If you make a mistake, use the segmentation tool's undo option. After you have pressed the cropping button, the segmentation tool will pause itself automatically for you to inspect the capture. If you would like to make another cut, you need to press the pause button to reengage with the tool. Once you're satisfied, click the check mark icon to confirm. CloudCompare will create a new layer for the cropped version, and you can disable or delete the unwanted data. Thanks for watching. If you're interested to learn more, you can check out the Learn page on Polycam, or you can leave a comment if you found this tutorial helpful.