Method for recognizing oscillations and/or deflection movements of an infrastructure component

1 . A method for recognizing oscillations and/or deflection movements of an infrastructure component relative to surroundings of the infrastructure component, the infrastructure component accommodating at least one LIDAR sensor and including a transmission tower or a sign gantry, the method comprising the following method steps:

a) detecting static and/or geometric features of a first point cloud during initial installation, using the at least one LIDAR sensor of the infrastructure component;

b) optionally segmenting the static features of the first point cloud from step a) using an algorithm, and generating a first segmented point cloud;

c) further detecting static and/or geometric features of a second point cloud at a later point in time than the initial installation, using the at least one LIDAR sensor of the infrastructure component;

d) segmenting the static features of the second point cloud from step c) using an algorithm, and generating a second segmented point cloud;

e) superimposing the first point cloud obtained in step a) or the first segmented point cloud obtained in step b) with the second segmented point cloud obtained in step d), using an algorithm;

f) when the first point cloud or the first segmented point cloud is superimposed with the second segmented point cloud, transformation parameters obtained using the algorithm correspond to values of a local oscillation of the infrastructure component and of the at least one LIDAR sensor accommodated therein; and

g) ascertaining an absolute oscillation of the infrastructure component.

2 . The method as recited in claim 1 , wherein the first point cloud includes the ground and/or roadway boundaries and/or pillars and/or road signs and/or traffic lights and/or sign gantries and/or roadway boundaries as the geometric and/or static features of the surroundings.

3 . The method as recited in claim 1 , wherein in step f) for superimposing the first point cloud or the first segmented point cloud with the second segmented point cloud, an algorithm is selected from a group including:

an iterative closest point algorithm,

a normal distribution transform (NDT) algorithm,

a neural deformation pyramid (NDP) algorithm,

an iterative dual correspondence (IDC),

a probabilistic iterative correspondence (pIC),

Gaussian fields,

point-based probabilistic registration,

quadratic patches,

likelihood field matching,

conditional random fields (CRFs),

branch-and-bound registration.

4 . The method as recited in claim 1 , wherein in step b), the segmentation of the static features of the first point cloud is carried out based on geometric objects including lines and/or curves and/or planes.

5 . The method as recited in claim 1 , wherein in step g), the absolute oscillation of the infrastructure component is ascertained based on its kinematic model.

6 . The method as recited in claim 5 , wherein the kinematic model includes a translatory portion and a rotatory portion.

7 . The method as recited in claim 1 , wherein the values of the local oscillations and the absolute oscillations of the infrastructure component and of the LIDAR sensors accommodated at the infrastructure component are added.

8 . The method as recited in claim 1 , wherein after an initial calibration and detection of the second point cloud using an NDT or ICP algorithm, a decalibration of the at least one LIDAR sensor is recognized and an autocalibration of the at least one LIDAR sensor is initiated.

9 . The method as recited in claim 8 , wherein a quality of the autocalibration of the at least one LIDAR sensor is assessed.

10 . A device for recognizing oscillations and/or deflection movements of an infrastructure component relative to surroundings of the infrastructure component, the infrastructure component accommodating at least one LIDAR sensor and including a transmission tower or a sign gantry, the device comprising:

an oscillation recognition module; and

the LIDAR sensor, the oscillation recognition module configured to:

a) detect static and/or geometric features of a first point cloud during initial installation, using the LIDAR sensor of the infrastructure component;

b) optionally segment the static features of the first point cloud from a) using an algorithm, and generate a first segmented point cloud;

c) further detect static and/or geometric features of a second point cloud at a later point in time than the initial installation, using the at least one LIDAR sensor of the infrastructure component;

d) segment the static features of the second point cloud from c) using an algorithm, and generate a second segmented point cloud;

e) superimpose the first point cloud obtained in a) or the first segmented point cloud obtained in b) with the second segmented point cloud obtained in d), using an algorithm;

f) when the first point cloud or the first segmented point cloud is superimposed with the second segmented point cloud, transformation parameters obtained using the algorithm correspond to values of a local oscillation of the infrastructure component and of the at least one LIDAR sensor accommodated therein; and

g) ascertain an absolute oscillation of the infrastructure component.

11 . An infrastructure component, comprising:

a transmission tower or a sign gantry; and

a device for recognizing oscillations and/or deflection movements of the infrastructure component relative to surroundings of the infrastructure component, the device including:

an oscillation recognition module, and

a LIDAR sensor accommodated at the infrastructure component,

wherein the oscillation recognition module configured to:

a) detect static and/or geometric features of a first point cloud during initial installation, using the LIDAR sensor of the infrastructure component;

b) optionally segment the static features of the first point cloud from a) using an algorithm, and generate a first segmented point cloud;

c) further detect static and/or geometric features of a second point cloud at a later point in time than the initial installation, using the at least one LIDAR sensor of the infrastructure component;

d) segment the static features of the second point cloud from c) using an algorithm, and generate a second segmented point cloud;

e) superimpose the first point cloud obtained in a) or the first segmented point cloud obtained in b) with the second segmented point cloud obtained in d), using an algorithm;

f) when the first point cloud or the first segmented point cloud is superimposed with the second segmented point cloud, transformation parameters obtained using the algorithm correspond to values of a local oscillation of the infrastructure component and of the at least one LIDAR sensor accommodated therein; and

g) ascertain an absolute oscillation of the infrastructure component.