A System Integrating Optical Music Recognition and Real-Time Playback with SCAMP by Yi Fu Chen, Ru Fang Guo, and Jung-Ching Chen

  • Published on 10/22/2025 by dora
OMR SCAMP

This project develops a web-based system that integrates Optical Music Recognition and real-time playback using YOLOv9 and SCAMP, enabling users to upload piano scores and instantly hear accurate, automatically rendered melodies.

This project presents an integrated system that combines Optical Music Recognition (OMR) with real-time playback using SCAMP. YOLOv9 is applied to both image segmentation and symbol recognition, achieving precision scores of 0.99 and 0.81, respectively. A preprocessing algorithm organizes recognition results, which are then translated into pitch and rhythm instructions for immediate playback and MP3 export via SCAMP. To enhance accessibility, a web-based interface allows users to upload piano score images and instantly hear the rendered melody. The system demonstrates strong potential for applications in music education, accessibility, and interactive learning.