Back to results
US 9,176,589 B2Granted
Gesture recognition apparatus and method of gesture recognition
AssigneeAAC Technology Pte, Ltd.
InventorsXiaojun Wang, Qi Tang
Life of this patent
Estimating the term…
Abstract
One embodiment of the invention discloses a gesture recognition apparatus including a left source configured to generate a left basic signal, a right source configured to generate a right basic signal, a detector configured to detect the left basic signal, the right basic signal and Doppler shift signals of the left and right basic signals after reflection by a hand, the detector disposed between the left source and the right source, and a processer configured to deal with signals from the detector and recognize a gesture. And a method of gesture recognition is also disclosed.
Claims17
A gesture recognition apparatus, comprising
- a left source configured to generate a left basic signal;
- a right source configured to generate a right basic signal;
- a detector configured to detect the left basic signal and the right basic signal without reflecting by a hand and Doppler shift signals of the left and right basic signals after reflection by the hand, the detector disposed between the left source and the right source;
- wherein, a difference of frequency between the left basic signal and the second basic signal is not less than 1000 Hz;
- a processer configured for dealing with signals from the detector with a Hamming window and FFT, obtaining a first frequency shift of the left basic signal and a second frequency shift of the right basic signal, normalizing the first frequency shift by dividing by an amplification of the left basic signal, calculating a sum of blue shift energy and a sum of red shift energy of the normalized first frequency shift, and obtaining a first frequency shift energy curve;
- wherein
- the processor is further configured for normalizing second frequency shift by dividing by an amplification of the right basic signal, calculating a sum of blue shift energy and a sum of red shift energy of the normalized second frequency shift signal, and obtaining a second frequency shift energy curve;
- obtaining a first two-value frequency shift energy curve and a second two-value frequency shift curve by setting thresholds;
- extracting a state sequence from the two two-value frequency shift energy curves;
- comparing the state sequence with a state sequence module board and recognizing the corresponding gesture;
- and outputting a corresponding operation.
8 more elements
A gesture recognition apparatus, comprising
- a first speaker configured to generate a left basic signal;
- a second speaker configured to generate a right basic signal;
- a microphone configured to detect the left basic signal and the right basic signal without reflecting by a hand and Doppler shift signals of the left and right basic signals after reflection by the hand, the microphone disposed between the left speaker and the right speaker;
- wherein, a difference of frequency between the left basic signal and the second basic signal is not less than 1000 Hz;
- a processer configured for dealing with signals from the detector with a Hamming window and FFT, obtaining a first frequency shift of the left basic signal and a second frequency shift of the right basic signal, normalizing the first frequency shift by dividing by an amplification of the left basic signal, calculating a sum of blue shift energy and a sum of red shift energy of the normalized first frequency shift, and obtaining a first frequency shift energy curve;
- wherein
- the processor is further configured for normalizing second frequency shift by dividing by an amplification of the right basic signal, calculating a sum of blue shift energy and a sum of red shift energy of the normalized second frequency shift signal, and obtaining a second frequency shift energy curve;
- obtaining a first two-value frequency shift energy curve and a second two-value frequency shift curve by setting thresholds;
- extracting a state sequence from the two two-value frequency shift energy curves;
- comparing the state sequence with a state sequence module board and recognizing the corresponding gesture;
- and outputting a corresponding operation.
8 more elements
method of gesture recognition, comprising the steps of:
- sending a left basic signal using a left source;
- sending a right basic signal using a right source;
- detecting the left basic signal, right basic signal and Doppler shift signals of the left and right basic signals after reflection by a hand using a detector;
- processing the signals from the detector with a Hamming window and FFT and obtaining a first frequency shift of the left basic signal and a second frequency shift of the right basic signal;
- normalizing the first frequency shift by dividing by an amplification of the left basic signal, calculating a sum of blue shift energy and a sum of red shift energy of the normalized first frequency shift, and obtaining a first frequency shift energy curve;
- normalizing second frequency shift by dividing by an amplification of the right basic signal, calculating a sum of blue shift energy and a sum of red shift energy of the normalized second frequency shift signal, and obtaining a second frequency shift energy curve;
- obtaining a first two-value frequency shift energy curve and a second two-value frequency shift curve by setting thresholds;
- extracting a state sequence from the two two-value frequency shift energy curves;
- comparing the state sequence with a state sequence module board and recognizing the corresponding gesture;
- outputting a corresponding operation.
7 more elements
Dependencies are inferred from the claim text.