Point Clouds In An Infrastructure Project Part One

For those of you that didn’t manage to get to our Point Clouds in Infrastructure events at our Southampton and Shipley offices I thought I’d give you a quick overview of what we covered. It’s too long for a blog post so I’ll split it up into nice chunks starting here with what we did in Recap.

If you want more information, contact the Graitec team and we’d be happy to talk to you about this or any other aspects of an infrastructure project.

What are Point Clouds?

Point Clouds are, put simply, lots of points in space. When I say lots, I mean up to billions of points to capture everything in the real world (hence the name, RECAP, or Reality Capture). Point Clouds capture everything they see, from buildings to bridges to passing people and cars to stones on the road and birds flying high above. Here lies the first challenge. You just need what you need. Not everything is relevant to your project.

There are 3 main ways to capture a point cloud, a land-based unit on a tripod, a mobile unit on a vehicle or from an airborne unit like a drone or an aircraft. You can either use a normal digital camera and capture images that can be processed through Recap Photo (this comes with Recap Pro) or with a laser scanner and process it through Recap Pro.

If you think about it, a point in space can have 3 sets of numbers. An X.Y & Z value. And an image can have 3 numbers for each pixel. A Red, Green & Blue value (RGB). These 2 sets of numbers can be interchanged. An XYZ value can be assigned to a picture’s pixel and an RGB code can be assigned to a point on an object. You will currently get better results with a laser scanner as it does not rely on an algorithm to interpolate the position of a pixel.

The resultant Point Cloud can be used to create a visualisation of the environment, a terrain model, or an object. And intern we can see the scene, get accurate measurements, communicate effectively without being on site and ultimately model our new intensions on an accurate survey.

What we covered

At the event we concentrated on 3 things we would like from a point cloud. A Terrain model (a surface), Linear features (road markings, Kerb lines etc) and Vertical features (Trees and lampposts etc). This was so we could make a Civil 3D model ready for designing a new roadway alignment or a widening of an existing road for example.

How we did this

We started with the raw Point Cloud data straight from the scanner in a .LAS format. Then using Recap Pro, we imported the data and the software auto registered it. That is to say it aligned all the files together into one scene. We then viewed and analysed the model ready for sharing and collaborating on the project.

There are different ways of viewing the data, such as RGB colours, Height or Intensity data themes:

Each theme has its uses. For example, you can see on the images above, Intensity highlights shiny objects like the white lines and signposts. The tarmac on the road is also easily identified in this case. Perfect for the type of job we are doing.

Images and videos can be exported.

Click on the image below to watch my Point Clouds in an Infrastructure Project video:

Recap Pro can also be used to clean up the data and reduce the size of the model to make it more manageable and relevant to the task in hand. Remember, Point Clouds catch everything so all departments can take out what they need from it.

We used the Limit box to reduce the point cloud to the area we are working in.

And then created regions and clipped out all the points we didn’t need.

Basically, we went from this:

To this:

                                                                                       To this:

 

 

 

 

 

 

 

 

This allows us to easily generate a surface for the road and extract linear features, such as the white the lines and the kerb edges.

In the next post I will show you how Infraworks helped us by automating a lot of our requirements and then ultimately into Civil 3D ready for modelling our new design.