Service: Geometry

This API is officially released and maintained.

Service offering on-demand geometric processing tasks.

Table 1. Available methods
Method	Request	Response
ComputeGroundAlignment	GeometryComputeGroundAlignmentRequest	GeometryComputeGroundAlignmentResponse
ComputeRegistrationRefinement	GeometryComputeRegistrationRefinementRequestStream	GeometryComputeRegistrationRefinementResponse

ComputeGroundAlignment

Computes the transformation aligning the lidar point cloud to the ground plane/gravity.

Request

Request for estimating alignment to ground/gravity based on imu-data and optional point cloud frame.

Table 2. Available fields in blickfeld.percept_toolkit.services.GeometryComputeGroundAlignmentRequest
Field	Type	Default	Unit
accelerometer_data	core_processing.data.Acceleration	-	-
accelerometer_data	accelerometer data is assumed to be expressed in the lidar coordinate frame
lidar_frame	optional core_processing.data.Frame	-	-
lidar_frame	if a lidar frame is provided, it is used to estimate the z-coordinate of the translation as well.
strategy	optional Strategy	-	-
strategy	Ground plane alignment estimation strategy.

Strategy

Ground plane alignment strategy options.

Table 3. Available values for blickfeld.percept_toolkit.services.GeometryComputeGroundAlignmentRequest.Strategy enum
Name	Value	Description
STRATEGY_UNSPECIFIED	0	Unspecified. Ground plane estimation will use STRATEGY_AUTO.
STRATEGY_AUTO	1	Automatic mode. Ground plane alignment will use a combination of plane segmentation & accelerometer.
STRATEGY_PLANE_SEGMENTATION	2	Point cloud only. Use the orientation and translation computed from the plane segmentation.
STRATEGY_ACCELEROMETER	3	Accelerometer only. Use the orientation from the accelerometer.

Response

The estimated alignment to the ground/gravity.

Table 4. Available fields in blickfeld.percept_toolkit.services.GeometryComputeGroundAlignmentResponse
Field	Type	Default	Unit
alignment	base.geometry.Transform	-	-
alignment	Estimated alignment of the sensor to the ground-plane/gravity. If no lidar frame is provided only roll, pitch of the rotation are estimated. Otherwise the z-component of the translation is additionally aligned with a flat ground. The transformation transforms the lidar point clouds such that the points are aligned with the ground.

ComputeRegistrationRefinement

Refines the point cloud registration based on provided initial transformations. If the method succeeds, the refined transformation is returned. Otherwise an exception will be thrown.

Request Stream

The request for transformation refinement based on 3d point clouds and an initial point cloud transformations.

Table 5. Available fields in blickfeld.percept_toolkit.services.GeometryComputeRegistrationRefinementRequest
Field	Type	Default	Unit
frame	core_processing.data.Frame	-	-
frame	lidar frame used for aligning the 3d point clouds
transform	base.geometry.Transform	-	-
transform	Transform: the transformation transforming a point in from the local lidar coordinate system into a 'combined' reference system the 'combined' reference system can also be a lidar.

Response

The refined lidar transformations.

Table 6. Available fields in blickfeld.percept_toolkit.services.GeometryComputeRegistrationRefinementResponse
Field	Type	Default	Unit
transformations	repeated base.geometry.Transform	-	-
transformations	Refined transformations. The array sequence is equal to the sequence in which the requests were received. The first element in the array corresponds to the lidar frame received in the first client stream message.