公开(公告)号：US11455503B2

公开(公告)日：2022-09-27

申请号：US17030700

申请日：2020-09-24

Applicant: Axis AB

Inventor： Andres Vigren ,  Aras Papadelis ,  Daniel Ståhl ,  Arnab Bhattacharjee

IPC: G06K9/62 ,  G06N3/04

Abstract: Apparatus are provided for generating an object classification for an object, the apparatus comprising an image sensor, a radar sensor, and a processing unit, the processing unit configured to perform the steps of: receiving image data for the object from the image sensor, operating an image-based object classifier on the image data to generate an image-based object classification, receiving radar data for the object from the radar sensor, operating a radar-based object classifier on the radar data to generate a radar-based object classification, selecting between the image-based object classification and the radar-based object classification to output as the object classification for the object, determining if a training condition is met by the radar-based object classification, and training the radar-based object classifier using the image-based object classification when the training condition is met by the radar-based object classification.

公开(公告)号：US12276722B2

公开(公告)日：2025-04-15

申请号：US17748223

申请日：2022-05-19

Applicant: Axis AB

Inventor： Anders Mannesson ,  André Nüsslein ,  Anton Sedin ,  Aras Papadelis ,  Daniel Ståhl ,  David Wadmark ,  Sebastian Heunisch ,  Stefan Adalbjörnsson

IPC: G01S13/88 ,  G01S13/72

Abstract: A method for classifying tracks in radar detections of a scene acquired by a stationary radar unit, comprises: acquiring radar detections of the scene using the static radar unit; feeding at least a portion of the radar detections into a tracker module for producing track-specific feature data indicating a specific track in the scene, feeding at least a portion of the radar detections into a scene model comprising information about scene-specific features aggregated over time, and information indicating areas in the scene with expected ghost target detections and areas with expected real target detections, wherein at least a subset of the scene-specific features is determined from the radar detections; classifying the specific track as belonging to a real target or to a ghost target by relating the specific track to a position in the scene model.