How Filtering is Used in 3D Smart Sensors

In Gocator, filters are used to post-process scan data along the X or Y axis to eliminate occlusions; remove noise, outliers, and edge effects; and “clean up” data before it is used by built-in measurement tools. The type of filter you choose depends on your specific application needs.

Gap Filling for Eliminating Occlusions

Gap Filling is used to fill in missing data points caused by occlusions. Gap Filling uses either the lowest values from the nearest neighbors or linear interpolation between neighboring values (depending on the Z difference between neighboring values), in a specified X or Y window.  

Gap Filling also fills gaps where no data is detected, which can be caused by the target’s surface reflectivity—for example dark or specular surface areas or actual gaps in the target surface. With this capability you can accurately scan challenging surface areas in applications such as electronic small parts inspection.

Gocator 2x00 Filters Gap Filling

Gocator’s easy-to-use Filters panel 

Median and Smoothing For Eliminating Outliers and Random Noise

The Median Filter substitutes the value of a data point with the median calculated within a specified window around the data point. Median is highly effective at random noise reduction, capable of completely discarding any outliers (very high or low measurements).

The Smoothing Filter is also used to reduce random noise in the scan data. Smoothing is faster than Median and works by substituting a data point value with the average value of that data point and its nearest neighbors within a specified window. Smoothing is most useful in high-speed applications and when scanning objects with lower noise levels.

Sample Results of Median Filter

Example of the Median Filter’s effectiveness at reducing random noise

Smart Feature: Part Edge Filtering for Reducing Edge Effects

Part scans sometimes contain noise around the edges of the target. This noise is usually caused by the sensor’s light being reflected off vertical sides, rounded corners, etc.

The Part Edge Filtering feature in Gocator is another way to reduce edge noise in order to produce more accurate and repeatable volume and area measurements, as well as to improve the positioning of relative measurement regions used for anchoring.

With Part Edge Filtering disabled:

Part Edge Filtering Disabled

With Part Edge Filtering enabled: 

Part Edge Filtering Enabled

Decimation for Increased Data Processing Speed

The Decimation Filter reduces the number of data points along the X or Y axis by choosing data points at the end of a specified window around the data point. For example, by setting X to 0.2, only points every 0.2 millimeters will be used.

Decimation increases Gocator’s speed by reducing the amount of data that measurement tools need to process—with a small tradeoff in resolution. This filter is useful in applications where processing speed takes precedent over high-resolution, for example in high-volume inline inspection of large parts.

Slope for Inspecting Rotating or Tilted Objects

Available on Gocator Single Point Profile Sensors, the Slope Filter extracts sharp changes within a profile in the presence of slow vibration, or other sources of incremental change such as positional variation. The Slope Filter is useful for inspecting defects on an object like a rotating or slightly off-center shaft or drum.