Radar data processing using neural network classifier and confidence metrics
A radar data processing device includes at least one analog-to-digital converter (ADC) configured to digitize a plurality of input signals, wherein each input signal includes radar chirp and radar chirp reflection information received at one of a plurality of receiver antennas. The radar data processing device also includes Fast Fourier Transform (FFT) logic configured to generate FFT output samples based on each digitized input signal, wherein at least some of the generated FFT output samples are across antenna FFT output samples associated with at least two of the plurality of receiver antennas. The radar data processing device also includes a processor configured to determine a plurality of object parameters based on at least some of the generated FFT output samples, wherein the processor uses a neural network classifier trained to provide a confidence metric for at least one of the plurality of object parameters.
1. A radar data processing device that comprises:
at least one analog-to-digital converter (ADC) configured to digitize a plurality of input signals, wherein each input signal includes radar chirp and radar chirp reflection information received at one of a plurality of receiver antennas;
Fast Fourier Transform (FFT) logic configured to generate FFT output samples based on each digitized input signal, wherein at least some of the generated FFT output samples are across antenna FFT output samples associated with at least two of the plurality of receiver antennas; and
a processor configured to detect the presence of at least one object in the generated FFT output samples relative to a constant false alarm rate (CFAR) detection threshold and determine a plurality of object parameters for the at least one detected object based on at least some of the generated FFT output samples, wherein the processor uses a neural network classifier trained to provide a confidence metric for at least one of the plurality of object parameters;
and wherein the processor is further configured to adjust the CFAR detection threshold responsive to the confidence metric for the at least one of the plurality of object parameters.
2. The device of claim 1 , wherein the neural network classifier comprises an artificial neural network (ANN) configured to provide a confidence metric for direction of arrival data obtained using at least some of the generated FFT output samples.
3. The device of claim 2 , wherein the direction of arrival data is a function of azimuth and elevation.
4. The device of claim 1 , wherein the neural network classifier is configured to only analyze FFT output samples and to provide a confidence metric for an object parameter associated with a previously detected object.
5. The device of claim 1 , wherein the neural network classifier is configured to analyze FFT output samples and to provide a confidence metric used for object detection.
6. The device of claim 1 , wherein the neural network classifier is trained to analyze FFT output samples and to provide a confidence metric used for object tracking.
7. The device of claim 1 , wherein the confidence metric is based on at least one of: strength of signals received by the at least two receiver antennas; a pattern of signals received by the at least two receiver antennas; a distribution of FFT output samples as a function of range and Doppler; and a distribution of FFT output samples as a function of elevation and azimuth.
8. The device of claim 1 , wherein the processor is configured to adjust the CFAR detection threshold by:
responsive to the confidence metric for the at least one of the plurality of object parameters exceeding a threshold, decreasing the CFAR detection threshold.
9. The device of claim 1 , wherein the processor is configured to adjust the CFAR detection threshold by:
determining an amount of adjustment of the CFAR detection threshold responsive to the confidence metric of the at least one of the plurality of object parameters.
10. An integrated circuit that comprises:
Fast Fourier Transform (FFT) logic configured to receive digitized input signals that include radar chirp and radar chirp reflection information received at a plurality of receiver antennas and to generate FFT output samples based on the digitized input signals, wherein at least some of the generated FFT output samples are across antenna FFT output samples associated with at least two of the plurality of receiver antennas; and
a processor configured to detect the presence of at least one object in the generated FFT output samples relative to a constant false alarm rate (CFAR) detection threshold and determine a plurality of object parameters for the at least one detected object based on at least some of the generated FFT output samples, wherein the processor uses a neural network classifier trained to provide a confidence metric for at least one of the plurality of object parameters;
and wherein the processor is further configured to adjust a constant false alarm rate (CFAR) detection threshold responsive to the confidence metric for the at least one of the plurality of object parameters.
11. The integrated circuit of claim 10 , wherein the neural network classifier comprises an artificial neural network (ANN) configured to provide a confidence metric for direction of arrival data obtained using at least some of the generated FFT output samples.
12. The integrated circuit of claim 11 , wherein the direction of arrival data is a function of azimuth and elevation.
13. The integrated circuit of claim 10 , wherein the neural network classifier is configured to only analyze FFT output samples and to provide a confidence metric for an object parameter associated with a previously detected object.
14. The integrated circuit of claim 10 , wherein the neural network classifier is configured to analyze FFT output samples and to provide a confidence metric used for object detection.
15. The integrated circuit of claim 10 , wherein the neural network classifier is configured to analyze FFT output samples and to provide a confidence metric used for object tracking.
16. The integrated circuit of claim 10 , wherein the processor is configured to adjust the CFAR detection threshold by:
responsive to the confidence metric for the at least one of the plurality of object parameters exceeding a threshold, decreasing the CFAR detection threshold.
17. The integrated circuit of claim 10 , wherein the processor is configured to adjust the CFAR detection threshold by:
determining an amount of adjustment of the CFAR detection threshold responsive to the confidence metric of the at least one of the plurality of object parameters.
18. A method that comprises:
digitizing a plurality of input signals, wherein each input signal includes radar chirp and radar chirp reflection information received at one of a plurality of receiver antennas;
generating FFT output samples based on each digitized input signal, wherein at least some of the generated FFT output samples are across antennas FFT output samples associated with at least two of the plurality of receiver antennas;
detecting the presence of at least one object in the generated FFT output samples relative to a constant false alarm rate (CFAR) detection threshold;
storing the generated FFT output samples;
determining a plurality of object parameters for the at least one detected object based on the stored FFT output samples;
providing a confidence metric for at least one of the plurality of object parameters based on at least some of the stored FFT output samples and using a neural network classifier; and
responsive to the confidence metric for the at least one of the plurality of object parameters, adjusting the CFAR detection threshold.
19. The method of claim 18 , wherein providing the confidence metric comprises using an artificial neural network (ANN) trained to provide a confidence metric for direction of arrival data obtained using at least some of the stored FFT output samples, and wherein the direction of arrival data is a function of azimuth and elevation.
20. The method of claim 18 , wherein providing the confidence metric comprises only analyzing FFT output samples and providing a confidence metric for an object parameter associated with a previously detected object.
21. The method of claim 18 , further comprising performing object detection based on the confidence metric.
22. The method of claim 18 , further comprising performing object tracking based on the confidence metric.
23. The method of claim 18 , wherein the step of adjusting the CFAR detection threshold comprises:
responsive to the confidence metric for the at least one of the plurality of object parameters exceeding a threshold, decreasing the CFAR detection threshold.
24. The method of claim 18 , wherein the step of adjusting the CFAR detection threshold further comprises:
determining an amount of adjustment of the CFAR detection threshold responsive to the confidence metric of the at least one of the plurality of object parameters.