Point Cloud Filter

Algorithm for filtering point clouds

Table 1. Available fields in blickfeld.percept_pipeline.config.PointCloudFilter
Field	Type	Default	Unit
radius_outlier	RadiusOutlier	-	-
radius_outlier	Use radius outlier filter for noise reduction on point clouds
adaptive_noise	AdaptiveNoise	-	-
adaptive_noise	Use adaptive noise filter to remove points based on their reflectivity

Radius Outlier

Filter points based on the number of neighbors in a certain radius

Table 2. Available fields in blickfeld.percept_pipeline.config.PointCloudFilter.RadiusOutlier
Field	Type	Default	Unit
min_neighbor_points	optional uint32	-	-
min_neighbor_points	Minimum number of neighbors a point has to have to be retained
neighbor_radius	optional double	-	-
neighbor_radius	All points within this radius are considered neighbors of a point

Adaptive Noise

Filter points based on a reflectivity threshold. The threshold is set to a configurable percentile of the reflectivities of the outliers from the radius outlier filter

Table 3. Available fields in blickfeld.percept_pipeline.config.PointCloudFilter.AdaptiveNoise
Field	Type	Default	Unit
max_reflectivity_threshold	optional uint32	-	-
max_reflectivity_threshold	Maximum value for the reflectivity threshold default: 1000
outlier_reflectivity_percentile	optional uint32	-	-
outlier_reflectivity_percentile	Percentile of the outlier reflectivities that is used as filtering threshold default: 85 (85th percentile)