Gesture recognition device and method using radar
A gesture recognition device and method using radar are proposed. The gesture recognition device includes: a signal receiving unit for receiving a radar signal reflected by a gesture of a user; a clutter removing unit for removing clutter from the signal received by the signal receiving unit; and a signal magnitude variance acquiring unit for acquiring the variance of a signal magnitude from a reflection signal from which the clutter has been removed. The proposed device and method have an advantage of enabling a gesture to be recognized with performance that is robust to changes in distance and direction between a user and a radar sensor.
1. A gesture recognition device using a radar, the gesture recognition device comprising:
a signal receiver processor configured to receive radar signals reflected by a gesture of a user;
a clutter remover processor configured to remove clutter from the signals received at the signal receiver processor;
a signal intensity variance acquirer processor configured to acquire a variance of signal intensity values from the reflected signals having the clutter removed therefrom;
a times of arrival (ToA) variance acquirer processor configured to acquire ToA's corresponding to maximum values of fast time signals forming the reflected signals having the clutter removed therefrom and configured to acquire a variance of the acquired ToA's corresponding to the maximum values;
a frequency acquirer processor configured to compute a frequency of the reflected signals having the clutter removed therefrom;
a reference database configured to store reference feature information of a predetermined plurality of reference gestures; and
a gesture recognition processor configured to determine which reference gesture the gesture of the user corresponds to by using the signal intensity variance, the maximum value ToA variance, the frequency, and the reference feature information stored in the reference database.
2. The gesture recognition device according to claim 1 , further comprising a fitting processor configured to determine whether or not the gesture is an intended gesture having intentional periodicity in the reflected signals having the clutter removed therefrom.
3. The gesture recognition device according to claim 1 , wherein the frequency acquirer processor determines whether the gesture of the user is a gesture with a large movement or a gesture with a small movement based on the maximum value ToA variance.
4. The gesture recognition device according to claim 3 , wherein the frequency acquirer processor performs a fast Fourier transform (FFT) on the reflected signals having the clutter removed therefrom to compute the frequency, if the gesture of the user is determined to be the gesture with the small movement.
5. The gesture recognition device according to claim 3 , wherein the frequency acquirer processor computes the frequency by subtracting an average of the ToA's corresponding to the maximum values from the ToA corresponding to the maximum value for each of the fast time signals and performing the FFT on resultant signals, if the gesture of the user is determined to be the gesture with the large movement.
6. The gesture recognition device according to claim 3 , wherein the fitting processor determines whether or not a signal has periodicity by way of a sinusoidal fitting for the gesture with the small movement and determines whether or not the signal has periodicity by way of an R-square fitting for the gesture with the large movement.
7. The gesture recognition device according to claim 1 , wherein the reference feature information comprises a signal intensity variance, a maximum value ToA variance, and K-means clustering information for frequency for each reference gesture.
8. The gesture recognition device according to claim 7 , wherein the gesture recognition processor recognizes a gesture by using the signal intensity variance, maximum value ToA variance, and frequency acquired from the gesture of the user and distance information from a cluster center in a K-means space for each of the reference gestures.
9. The gesture recognition device according to claim 1 , wherein the gesture recognition processor performs gesture recognition after normalizing the signal intensity variance, maximum value ToA variance, and frequency.
10. A gesture recognition method using a radar, the gesture recognition method comprising:
(a) receiving radar signals reflected by a gesture of a user;
(b) removing clutter from the signals received at said step (a);
(c) acquiring a signal intensity variance from the reflected signals having the clutter removed therefrom;
(d) acquiring times of arrival (ToA's) corresponding to maximum values of fast time signals forming the reflected signals having the clutter removed therefrom and acquiring a maximum value ToA variance defined as a variance of the acquired ToA's corresponding to the maximum values;
(e) computing and acquiring a frequency of the reflected signals having the clutter removed therefrom; and
(f) determining which reference gesture the gesture of the user corresponds to by using ‘reference feature information of a predetermined plurality of reference gestures and the signal intensity variance, the maximum value ToA variance, and the frequency.
11. The gesture recognition method according to claim 10 , wherein said step (e) comprises determining whether the gesture of the user is a gesture with a large movement or a gesture with a small movement based on the maximum value ToA variance.
12. The gesture recognition method according to claim 10 , wherein said step (e) comprises performing a fast Fourier transform (FFT) on the reflected signals having the clutter removed therefrom to compute the frequency, if the gesture of the user is determined to be a gesture with a small movement, and computing the frequency by subtracting an average of the ToA's corresponding to the maximum values from the ToA corresponding to the maximum value for each of the fast time signals and performing the FFT on the resultant signals, if the gesture of the user is determined to be a gesture with a large movement.
13. The gesture recognition method according to claim 10 , wherein the reference feature information comprises a signal intensity variance, a maximum value ToA variance, and K-means clustering information for frequency for each reference gesture.
14. The gesture recognition method according to claim 13 , wherein said step (f) comprises recognizing a gesture by using the signal intensity variance, maximum value ToA variance, and frequency acquired from the gesture of the user and distance information from a cluster center in a K-means space for each of the reference gestures.
15. The gesture recognition method according to claim 10 , further comprising normalizing the acquired signal intensity variance, maximum value ToA variance, and frequency.