Fig. 2
From: Efficient music identification using ORB descriptors of the spectrogram image
Proposed system. The figure shows the proposed system where the raw audio would be captured (by loading the audio signal or by microphone). The audio signal is then processed by converting the audio signal into a spectrogram image. The spectrogram image of the sample is compared to the database of spectrogram images of the songs in the database. This is done using an image processing techniques to match the features of the spectrogram image. After the resulting audio fingerprint comparison is done, the corresponding song information for the highest match is returned