This study presents the development of an innovative music learning system that integrates artificial intelligence-based image recognition technology, employing interactive musical storybooks and intelligent musical blocks as core pedagogical tools to establish a novel digital learning paradigm for musical literacy. Using a quasi-experimental design, 27 participants aged 7–12 years were randomly assigned to experimental and control groups. The system integrates Robflow’s image recognition technology with the YOLOv11 deep learning model, achieving 99.5% mean precision in musical note recognition tasks. Analysis using the Achievement Emotions Questionnaire (AEQ) revealed that the experimental group exhibited significantly higher positive emotions (enjoyment, hope, and pride) while demonstrating notable reductions in negative emotions (boredom, shame, and hopelessness). From a human-computer interaction perspective, the system employs CNC precision machining and epoxy resin filling techniques to develop transparent musical blocks that blend aesthetics with functionality. The research validates the feasibility and effectiveness of AI-assisted music education in fostering musical literacy, providing new directions for future digital interactive learning development.