Object classification for vehicle radar systems

Patent metadata

Field	Value
Publication number	US-9664779-B2
Application number	US-201414323415-A
Country	US
Kind code	B2
Filing date	Jul 3, 2014
Priority date	Jul 3, 2014
Publication date	May 30, 2017
Grant date	May 30, 2017

Abstract

Methods and systems are provided for object classification for a radar system of a vehicle. The radar system includes a transmitter that transmits radar signals and a receiver that receives return radar signals after the transmitted radar signals are deflected from an object proximate the vehicle. A processor is coupled the receiver, and is configured to: obtain spectrogram data from a plurality of spectrograms pertaining to the object based on the received radar signals; aggregate the spectrogram data from each of the plurality of spectrograms into a computer vision model; and classify the object based on the aggregation of the spectrogram data from each of the plurality of spectrograms into the computer vision model.

First claim

What is claimed is:

1. A method for object classification for a radar system of a vehicle, the method comprising the steps of:
obtaining, via a processor, spectrogram data from a plurality of spectrograms pertaining to an object proximate the vehicle based on received radar signals for the radar system;
aggregating, via the processor, the spectrogram data from each of the plurality of spectrograms into a clustering computer vision model;
classifying, via the processor, the object based on the aggregation of the spectrogram data from each of the plurality of spectrograms into the clustering computer vision model
extracting, via the processor, features from each of the plurality of spectrograms;
performing, via the processor, a vector quantization of the features; and
incorporating, via the processor, the vector quantization into the clustering computer vision model.

2. The method of claim 1 , further comprising:
identifying a plurality of three dimensional spatial regions proximate the object, wherein the obtaining spectrogram data comprises obtaining spectrogram data from the plurality of spectrograms, each of the plurality of spectrograms pertaining to a corresponding one of the plurality of three dimensional spatial regions.

3. The method of claim 1 , wherein the incorporating the vector quantization comprises:
representing the features as a plurality of histograms of clusters using the vector quantization and a bag of words computer vision model.

4. The method of claim 1 , wherein the aggregating the spectrogram data comprises comparing energy intensities of the spectrogram data from the plurality of spectrograms via calculated sums of values from each of the plurality of spectrograms using the computer vision model.

5. The method of claim 4 , wherein the aggregating is performed using a modified regression tree computer vision model.

6. A radar control system for a vehicle, the radar control system comprising:
a transmitter configured to transmit radar signals;
a receiver configured to receive return radar signals after the transmitted radar signals are deflected from an object proximate the vehicle; and
a processor coupled the receiver and configured to:
obtain spectrogram data from a plurality of spectrograms pertaining to the object based on the received radar signals;
aggregate the spectrogram data from each of the plurality of spectrograms into a clustering computer vision model;
classify the object based on the aggregation of the spectrogram data from each of the plurality of spectrograms into the clustering computer vision model;
extract features from each of the plurality of spectrograms;
perform a vector quantization of the features; and
incorporate the vector quantization into the clustering computer vision model.

7. The radar control system of claim 6 , wherein the processor is further configured to:
identify a plurality of three dimensional spatial regions proximate the object, wherein each of the plurality of spectrograms pertains to a corresponding one of the plurality of three dimensional spatial regions.

8. The radar control system of claim 6 , wherein the processor is further configured to represent the features as a plurality of histograms of clusters using the vector quantization and a bag of words computer vision model.

9. The radar control system of claim 6 , wherein the processor is further configured to compare energy intensities of the spectrogram data from the plurality of spectrograms via calculated sums of values from each of the plurality of spectrograms using the clustering computer vision model.

10. The radar control system of claim 9 , wherein the clustering computer vision model comprises a modified regression tree computer vision model.

11. A computer system for object classification for a radar system of a vehicle, the computer system comprising:
a non-transitory, computer readable storage medium storing a program, the program configured to at least facilitate:
obtaining spectrogram data from a plurality of spectrograms pertaining to the object based on received radar signals for the radar system;
aggregating the spectrogram data from each of the plurality of spectrograms into a clustering computer vision model;
classifying the object based on the aggregation of the spectrogram data from each of the plurality of spectrograms into the clustering computer vision model;
extracting features from each of the plurality of spectrograms;
performing a vector quantization of the features; and
incorporating the vector quantization into the clustering computer vision model.

12. The computer system of claim 11 , wherein the program is further configured to at least facilitate:
identifying a plurality of three dimensional spatial regions proximate the object, wherein each of the plurality of spectrograms pertains to a corresponding one of the plurality of three dimensional spatial regions.

13. The computer system of claim 11 , wherein the program is further configured to at least facilitate representing the features as a plurality of histograms of clusters using the vector quantization and a bag of words computer vision model.

14. The computer system of claim 11 , wherein the program is further configured to at least facilitate comparing energy intensities of the spectrogram data from the plurality of spectrograms via calculated sums of values from each of the plurality of spectrograms using a modified regression tree computer vision model.

Assignees

• Gm Global Tech Operations Llc

Inventors

• Stainvas Olshansky Inna
• Bilik Igal

Classifications

• G01S13/931 Primary

  of land vehicles · CPC title

• G01S7/415 Primary

  Identification of targets based on measurements of movement associated with the target · CPC title

• G01S7/41 Primary

  using analysis of echo signal for target characterisation; Target signature; Target cross-section · CPC title

• G01S7/417

  involving the use of neural networks · CPC title

• G01S13/88

  Radar or analogous systems specially adapted for specific applications (electromagnetic prospecting or detecting of objects, e.g. near-field detection, G01V3/00) · CPC title