Gesture recognition using frequency modulated continuous wave (FMCW) radar with low angle resolution

• US 9,817,109 B2
• Filed: 02/27/2015
• Issued: 11/14/2017
• Est. Priority Date: 02/27/2015
• Status: Active Grant

• Current Assignee: Texas Instruments, Inc.
• Original Assignee: Texas Instruments, Inc.
• Inventors: Saboo, Krishnakant Vijay, Rao, Sandeep
• Primary Examiner(s): Brainard, Timothy A
• Application Number:
  US14/633,684
• Publication Number:
  US 20160252607A1
• Time in Patent Office: 991 Days
• Field of Search: 342107
• CPC Class Codes: G01S 13/02 Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
• CPC Class Codes: G01S 13/58 Velocity or trajectory determination systems; Sense-of-movement determination systems
• CPC Class Codes: G01S 13/584 adapted for simultaneous range and velocity measurements
• CPC Class Codes: G01S 7/356 involving particularities of FFT processing
• CPC Class Codes: G01S 7/415 Identification of targets based on measurements of movement associated with the target
• CPC Class Codes: G06F 3/011 Arrangements for interaction with the human body, e.g. for user immersion in virtual reality blind teaching G09B21/00
• CPC Class Codes: G06F 3/017 Gesture based interaction, e.g. based on a set of recognized hand gestures interaction based on gestures traced on a digitiser G06F3/04883

Abstract

A method for operating a frequency modulated continuous wave (FMCW) radar system is provided that includes generating digital intermediate frequency (IF) signals from radio frequency signals received by a small receive antenna array in the FMCW radar system and processing the digital IF signals to determine whether or not a gesture was performed.

Claims

1. A method for operating a frequency modulated continuous wave (FMCW) radar system, the method comprising:

generating digital intermediate frequency (IF) signals from radio frequency signals received by a small receive antenna array in the FMCW radar system;

computing a sequence of weighted velocities based on the digital IF signals;

detecting at least one peak in the sequence of weighted velocities; and

performing gesture classification based on parameters of the at least one peak to determine whether or not the at least one peak is indicative of a gesture.

2. The method of claim 1, in which the small receive antenna array consists of one receive antenna.

3. The method of claim 1, including filtering the sequence of weighted velocities prior to the detecting at least one peak.

4. The method of claim 1, in which processing the digital IF signals includes computing a sequence of weighted angles based on the digital IF signals, and in which performing gesture classification includes performing gesture classification based on the parameters and on weighted angles corresponding to the at least one peak.

5. The method of claim 1, in which the at least one peak includes two peaks and performing gesture classification includes recognizing the two peaks as a gesture and determining a direction of the gesture based on a sign of a weighted angle corresponding to a first peak of the two peaks and a sign of a weighted angle corresponding to a second peak of the two peaks.

6. The method of claim 1, in which computing a sequence of weighted velocities includes:

receiving a plurality of digital IF signals, one from each antenna of the antenna array;

computing a two dimension (2D) FFT for each digital IF signal of the plurality of digital IF signals to generate a 2D grid for each IF signal;

averaging amplitudes of corresponding bins of the 2D grids to generate an averaged 2D grid; and

computing a weighted velocity based on an expected location of a gesture in the averaged 2D grid.

7. The method of claim 6, in which computing a sequence of weighted velocities includes estimating the expected location.

8. The method of claim 1, including:

detecting a position cue while the FMCW radar system is operating in a normal mode having a first frame time and a first duty cycle;

transitioning the FMCW radar system to a gesture recognition mode responsive to the position cue, the gesture recognition mode having a second frame time and a second duty cycle; and

operating the FMCW radar system in the gesture recognition mode in which the generating digital IF signals and the processing the IF signals are performed.

9. The method of claim 8, including:

receiving an indication to terminate the gesture recognition mode; and

transitioning the FMCW radar to the normal mode.

10. A frequency modulated continuous wave (FMCW) radar system comprising:

a small receive antenna array;

a receive channel for each receive antenna in the receive antenna array, each receive channel coupled to a respective receive antenna to generate a digital intermediate frequency (IF) signal from a radio frequency signal received by the receive antenna array;

a digital front end coupled to receive the digital IF signals and produce digital baseband signals; and

a processor coupled to each receive channel to process the digital baseband signals, compute a sequence of weighted velocities from the baseband signals, and detect at least one peak in the sequence of weighted velocities to determine whether or not a gesture was performed.

11. The FMCW radar system of claim 10, in which the small receive antenna array consists of one receive antenna.

12. The FMCW radar system of claim 10, wherein the receive channel is configured to filter an intermediate frequency and produce the digital IF signals.

13. The FMCW radar system of claim 10, wherein the processor is configured to compute a sequence of weighted angles based on the baseband signals, and perform gesture classification based on parameters of the at least one peak and on weighted angles corresponding to the at least one peak.

14. The FMCW radar system of claim 13, in which the at least one peak includes two peaks and performing gesture classification includes recognizing the two peaks as a gesture and determining a direction of the gesture based on a sign of a weighted angle corresponding to a first peak of the two peaks and a sign of a weighted angle corresponding to a second peak of the two peaks.

15. The FMCW radar system of claim 10, in which computing a sequence of weighted velocities includes:

computing a two dimension (2D) FFT for each receive channel to generate a 2D grid receive channel;

averaging amplitudes of corresponding bins of the 2D grids to generate an averaged 2D grid; and

computing a weighted velocity based on an expected location of a gesture in the averaged 2D grid.

16. The FMCW radar system of claim 15, in which computing a sequence of weighted velocities includes estimating the expected location.

17. The FMCW radar system of claim 10, in which the FMCW radar system is configured to operate in a normal mode and a gesture recognition mode, the normal mode having a first frame time and a first duty cycle and the gesture recognition mode having a second frame time and a second duty cycle, in which the processor causes the FMCW radar system to transition from normal mode to gesture recognition mode responsive to a position cue detected in digital IF signals received from the receive channels while the FMCW radar system is operating in normal mode.

18. The FMCW radar system of claim 17, in which the processor causes the FMCW radar system to transition from gesture recognition mode to normal mode responsive to receiving an indication to terminate gesture recognition mode.

19. A method for operating a frequency modulated continuous wave (FMCW) radar system, the method comprising:

generating digital baseband signals from radio frequency signals received by an antenna array of the FMCW radar system;

computing a sequence of weighted velocities from the digital baseband signals;

detecting at least one peak in the sequence of weighted velocities; and

performing gesture classification based on parameters of the at least one peak to determine whether or not the at least one peak is indicative of a gesture.

20. The method of claim 19, comprising computing a sequence of weighted angles from the digital baseband, wherein the gesture classification is based on the parameters and on a weighted angle of to the at least one peak.