Sparsity Aware Dynamic Gesture Classification Using Dual-band Radar

Abstract

In this paper, we aim to recognize dynamic hand gestures by analyzing the sparse micro-Doppler radar signatures collected by dual-band radar sensors. The radar echoes are firstly mapped into the time-frequency domain through the Gaussian-windowed Fourier dictionary at each radar sensor. Then the sparse time-frequency features are extracted via the orthogonal matching pursuit (OMP) algorithm. Finally, the sparse time-frequency features extracted at dual-band radar sensors are fused and inputted into the modified-Hausdorff-distance-based nearest neighbor (NN) classifier to achieve the dynamic hand gesture classification. The experimental results based on the measured data demonstrate that 1) the classification accuracy using dual-band radar sensors is higher than that using only single band radar sensor; 2) the classification accuracy can be improved as the percentage of training data is increased.