Gaussian Smoothing

The Gaussian filter performs smoothing of points within a small local region based on 3D distances. This can be used to suppress sparse noise, correct outliers, and align points to a grid. Sigma parameter defines the aggressiveness of the filter. The higher the sigma, the more aggressive smoothing. Note that the Gaussian filter can, in many cases, correct outliers, but high sigma will also smooth out edges.

Gaussian filter adds the following processing time per 3D capture:

When do you use the Gaussian smoothing?

Gaussian smoothing is useful when you want to reduce absolute noise in a point cloud before further processing. Lower noise levels can improve the robustness and stability of certain vision algorithms, particularly in cases where noise causes false detections or unstable results. By reducing high frequency variations, Gaussian smoothing makes the data more consistent overall. While the relative variation between regions remains the same, the absolute noise level is reduced, which can help algorithms produce more stable outputs.

Gaussian smoothing is especially beneficial when evaluating large scale geometric properties, such as plane flatness. In these cases, suppressing outliers and small local variations can prevent the algorithm from producing overly pessimistic measurements. However, smoothing also reduces fine detail. If an algorithm relies on sharp edges or small features, excessive smoothing may reduce detection accuracy. Therefore, the amount of smoothing should be chosen carefully based on the application.

Version History

SDK	Changes
2.9.0	Reduce filter processing time and improve edge preservation capabilities.
1.3.0	Gaussian filter API is added.