Reprodução de acordes musicais com mão o mecânica em um teclado a partir de um analisador de espectro sonoro
This work presents the elaboration of two methods for the recognition of musical chords (chords: Dó, Ré, Mi, Fá, Sol, Lá and Si), in which one of them will be used in a system capable of reproducing the chord analyzed in a musical keyboard with the aid of a mechanical hand. One of the methods is com...
| Autor principal: | Moraes, Hercules Vinicius Tabolka de |
|---|---|
| Formato: | Trabalho de Conclusão de Curso (Graduação) |
| Idioma: | Português |
| Publicado em: |
Universidade Tecnológica Federal do Paraná
2020
|
| Assuntos: | |
| Acesso em linha: |
http://repositorio.utfpr.edu.br/jspui/handle/1/14932 |
| Tags: |
Adicionar Tag
Sem tags, seja o primeiro a adicionar uma tag!
|
| Resumo: |
This work presents the elaboration of two methods for the recognition of musical chords (chords: Dó, Ré, Mi, Fá, Sol, Lá and Si), in which one of them will be used in a system capable of reproducing the chord analyzed in a musical keyboard with the aid of a mechanical hand. One of the methods is completely innovative and is based on frequency scanning. Each frequency range of the chord under analysis is analyzed so that a characteristic vector of 23 positions is formed and compared to 7 base vectors, where each base vector, which also has 23 positions, corresponds to a reference chord; the other method is based on pattern recognition, in which the main idea is to classify the analyzed object (chord) into one of the pre-established classes (base chords) through a distance measure known as Euclidean distance, and a classifier of the type supervised. The system can be divided into three parts: the first part consists of a classification algorithm whose purpose is to make the spectral recognition of the musical signal through the Fast Fourier Transform and compare it with reference musical chords; the second part of the system is responsible for making the connection between the recognition system and the object that will perform the musical chords on the music keyboard: a data acquisition board.The NI 6009 data acquisition board has 12 digital outputs and each output will be responsible for triggering a mechanical finger (with the help of a drive of activation). The chord recognition routine will command which outputs should be triggered so that the mechanical hand reproduces the musical chord recognized by the system; the last stage of the system is an electromechanical device capable of playing musical chords on a keyboard: a mechanical hand. An octave of a musical keyboard has 12 notes, so the mechanical hand has 12 fingers, each one of which is responsible for playing a musical note. Underneath each finger has a fixed ferromagnetic material and an iron bar surrounded by a spool of 1000 turns. When the coil is energized, the iron bar becomes an electromagnet and attracts the ferromagnetic material fixed on the finger, which produces a vertical movement (top to bottom) reproducing the note on the musical keyboard. At the end of the tests, the frequency scanning method showed a 64% efficiency in chord recognition, whereas the pattern recognition method, which was used in the system implementation, recognized the chords with an efficiency of 100%. In the end, the mechanical hand reproduced all the chords analyzed in a correct and fast way, meeting the objectives of the work. |
|---|