Musical Note Detection

This is a repository for the course project of COMPI302: Computer Vision.

Author: Yueqian Lin

Instructor: Prof. Matthias Schroeter, Ph.D.

How to Run

1. Put the image to be processed in the same directory as main.py.
2. Run python main.py test.png --temp_folder 'temp_folder' --output_dir 'output' --threshold 1--origin origin.
3. The result will be saved as result.png in the same directory. The pitches and durations of the notes will be printed in the terminal and saved as result.txt in the output directory.

File Structure

.
├── README.md
├── combine.py
├── find_pitches.py
├── main.py
├── pre_process.py
├── slides.pdf
├── symbol
│   ├── accidental
│   │   ├── flat.png
│   │   └── sharp.png
│   ├── clef
│   │   ├── bass_clef.png
│   │   ├── bass_clef1.png
│   │   ├── treble_clef.png
│   │   └── treble_clef1.png
│   ├── notehead
│   │   ├── half0.tif
│   │   ├── half1.tif
│   │   ├── quarter0.tif
│   │   ├── quarter1.tif
│   │   ├── quarter2.tif
│   │   ├── quarter3.tif
│   │   ├── whole0.png
│   │   └── whole1.png
│   ├── rest
│   │   └── rest_quarter.png
│   └── time
│       ├── 24.tif
│       ├── 34.tif
│       ├── 44.tif
│       └── 68.tif
└── test.png

File Description

• main.py: The main program.
• pre_process.py: Pre-process the image into batches of single staffs.
• find_pitches.py: Find the pitches and durations of the notes for each staff.
• combine.py: Combine the batches of single staffs into a whole piece of music.