Granted Patent B2

US 12,562,063 · App. 17/757,727 · Granted Feb 24, 2026

Inventors: Martin Pfitzer (Lindau, DE); Stefan Heinrich (Achern, DE); Manuel du Bois (Bad Freienwalde, DE); Felix Hachfeld (Lindau, DE); Elias Strigel (Wangen, DE)

Assignee: CONTINENTAL AUTONOMOUS MOBILITY GERMANY GMBH

G08G1/166G06V20/58G06V40/103G06V40/23G06V10/751

Abstract

Claims (25)

1 . A method for recognizing road users, in particular persons in road traffic, comprising:

recording at least one ambient image or a sequence of ambient images by an optical sensor of a vehicle;

detecting a person in the at least one recorded ambient image or the sequence of recorded ambient images;

determining a pose of the detected person;

determining a presence of a means of transport of the person, wherein the presence of the means of transport is determined by considering a time sequence of pose states of the detected person or by considering individual images of the at least one recorded ambient image or the sequence of recorded ambient images, wherein the means of transport is classified by a classifier, the classifier classifies the means of transport on the basis of the pose of the detected person and leg movements corresponding to the classification;

assigning a movement profile to the person based on the determined presence of the means of transport of the person and the classification of the means of transport;

determining a criticality in a probability of the person crossing a road based on the determined presence of the means of transport; and

reducing an intervention threshold value of an emergency brake assistant system when a predetermined criticality threshold of the person crossing the road is met.

2 . The method according to claim 1 , wherein

the assigned movement profile comprises at least one of a maximum speed, degrees of freedom of the person or a direction of movement.

3 . The method according to claim 1 , further comprising at least partly controlling, by an actuator of the vehicle, the vehicle based in part upon the movement profile.

4 . The method according to claim 1 , further comprising storing a distinct movement profile for each potential classification of a plurality of classifications of different means of transport.

5 . The method according to claim 1 , wherein the presence of the means of transport is determined based upon the pose states of the detected person being repeatable within a period of time.

6 . A system for recognizing road users, in particular persons in road traffic, comprising:

at least one optical sensor for recording an ambient image or a sequence of ambient images as well as a data processor which is configured to:

analyze the recorded ambient image or the sequence of recorded ambient images and to carry out person recognition;

carry out pose recognition of the person;

determine a presence of a means of transport of the person in the recorded ambient image or the sequence of recorded ambient images, wherein the presence of the means of transport is determined by considering a time sequence of pose states of the detected person or by considering individual images of the at least one recorded ambient image or the sequence of recorded ambient images, wherein the means of transport is classified by a classifier, the classifier classifies the means of transport on the basis of the pose of the detected person and leg movements corresponding to the classification;

assign a movement profile to the person based on the determined presence of the means of transport and the classification of the means of transport;

determine a criticality in a probability of the person crossing a road based on the determined presence of the means of transport; and

reduce an intervention threshold value of an emergency brake assistant system when a predetermined criticality threshold of the person crossing the road is met.

7 . The system according to claim 6 , wherein the assigned movement profile comprises a maximum speed, degrees of freedom of the person and/or a direction of movement.

8 . The system according to claim 6 , further comprising at least one vehicle actuator coupled to the data processor, the actuator at least partly controlling a vehicle based upon the movement profile.

9 . The system according to claim 6 , further comprising non-transitory memory coupled to the data processor and having stored therein a distinct movement profile for each potential classification of a plurality of classifications of different means of transport.

10 . The system according to claim 6 , wherein the presence of the means of transport is determined based upon the pose states of the detected person being repeatable within a period of time.

Priority Claims (1)

Continuity (1)

References Cited (37)