An Introduction to Control Points

What are Control Points?

Adding control points during a scan with a NavVis device improves point cloud accuracy, by supplying additional, external information about the device’s real-world position, automatically geo-referencing and aligning datasets.

This helps specifically in environments where multiple loop closures are not possible and longer distances need to be scanned.

We have demonstrated this in several scans and projects published in our whitepapers.

Why Does Drift Error Occur?

NavVis scanning technology is based on SLAM algorithms that use the information from the sensors to incrementally build a 3D model of a building, you can find more information here. This is a powerful and flexible approach, but like many scanning technologies, it works by chaining measurement after measurement. This chain of measurements is called the trajectory estimate. You can think of it as a trail of points describing where the device was, relative to the dataset origin, and at which point in time.

Even though each individual measurement has very high accuracy, chaining many of them together means that, over time, there can be a drift error in the device's position estimate. This can show up as distortions in the mapping results, such as a slight bending of long corridors compared to the true building geometry.

How To Reduce Drift

Loop closures are one way to reduce drift. Our algorithms recognize previously visited places and use that data to adjust the trajectory estimate. However, loop closures may not be feasible in every building layout, or you may need to achieve a higher accuracy to a higher degree of certainty. This is where measured control points come in.

Measured Control Points

Shown here is an example of how measured control points help you create a accurate map.

This graphic represents a building:

Place control points in the building and measure them with a total station.

The origin of this coordinate system is not the same as the origin of the dataset. The person measuring the control points chooses one origin for the entire site.

Measured control points added to a dataset during mapping then act as constraints on the trajectory estimate when the coordinate information is passed along to our algorithms during processing.

Compare the trajectory estimate to the real coordinates of the control points.

The post-processing algorithms undistort the datasets using the information from the control points.

The algorithms also extract alignment information based on the positions of the control points and datasets.

Note: During the time when SLAM is lost, adding control points is not possible.