A fast Gabor spectrogram

The Gabor expansion based spectrogram (commonly known as the Gabor spectrogram or the time-frequency distribution series) has previously been applied in many areas. In many applications, such as biomedical engineering, radar image processing, and seismology, it has been found to be a good alternative for the STFT-based spectrogram and Wigner-Ville distribution. Although the idea of the Gabor spectrogram is rather straightforward, the implementation has been quite involved. In this paper, we present a new algorithm for computing the Gabor spectrogram. The method presented is not only simple, but also fast. According to this new approach, the Gabor spectrogram is nothing more than separated two-dimensional interpolation filters. Consequently, it could be efficiently computed in the Fourier transform domain or implemented in off-the-shelf chips.