Back to results
US 2026/0110773 A1Published application
Object detection in cluttered radar data
AssigneeAxis AB
InventorsSanthosh Nadig, Anders Mannesson
Life of this patent
Estimating the term…
Loading patent…
Loading patent…
AssigneeAxis AB
InventorsSanthosh Nadig, Anders Mannesson
Estimating the term…
Object detection is addressed through a structured method that emphasizes thresholding and signal regularization. Radar measurements of a scene provide intensity data distributed across points in either a range-Doppler or range-angle plane. A one-dimensional subset of the selected plane is parametrized to establish a simplified domain for analysis. From this subset, intensity values falling below a predefined clutter threshold are extracted, isolating potentially useful information from the clutter background. These extracted values are then regularized with respect to the parameter, producing a signal mask that captures consistent variations while suppressing noise fluctuations. A detection threshold is derived from the signal mask and subsequently applied to the radar data of the subset to determine the presence of objects. This approach enhances robustness in cluttered environments by dynamically adapting the detection threshold, improving the accuracy and reliability of radar-based object identification.
A method of detecting objects in radar data, the method comprising:
A signal processing device comprising processing circuitry configured to detect objects in radar data by performing a method comprising:
A computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out a method comprising:
Dependencies are inferred from the claim text.