Real-time Guitar Chord Estimation by Stereo Cameras for Supporting Guitarists

Learning to play the guitar usually involves tedious lessons in fingering positions for the left hand. It is difficult for beginners to recognize by themselves whether they are accurately positioning their fingers on the string to make guitar chords. This paper proposes a novel approach for recognizing chords played on a guitar. In real time, it estimates whether a beginner is holding the guitar correctly. This system detects the position of a player's fingers in 3D and calculates the guitar chord which the player's left hand is producing. Stereo cameras are used to compute the 3D positions of fingers using the triangulation method. ARTag (Augmented Reality Tag) is utilized to calculate the extrinsic parameters in each frame, so that our system will allow users to move the guitar while they are playing. The experimental results have revealed that the proposed system is effective even when used in real time.