Automated Music Sheet to MIDI Conversion and Spectrogram Analysis for Multi-Instrument and Multi-Genre Composition

Automated music transcription has emerged as a crucial field in Optical Music Recognition (OMR) and digital signal processing, enabling seamless conversion of sheet music into machine-readable formats. This paper presents a comprehensive system for automated music sheet-to-MIDI conversion and spectrogram analysis, leveraging advanced image processing techniques, deep learning models, and MIDI synthesis algorithms. The proposed framework first processes scanned sheet music by detecting and extracting musical symbols using contour analysis, morphological operations, and the Hough Transform. A deep learning-based approach is employed for note head recognition, pitch estimation, and rhythm extraction, ensuring accurate musical interpretation. The extracted notes are converted into MIDI format, which is further synthesized into WAV audio for playback and spectrogram visualization. The system incorporates a user-defined playback duration feature, optimizing real-time applications for composers, researchers, and musicians. The generated spectrogram provides a time-frequency representation, facilitating detailed harmonic analysis. This frame- work integrates computer vision, artificial intelligence, and digital signal processing to enhance music transcription, performance evaluation, and AI-assisted composition.