Music education in colleges and universities is transforming towards digitalization and intelligence, and piano teaching, as the core curriculum of music education, faces the demand for technological innovation. Traditional piano teaching relies on teachers’ subjective judgment in terms of hand shape correction and technique training, and lacks objective quantitative standards. This study constructs a piano technique training optimization system based on computer vision technology, aiming to improve the scientificity and accuracy of piano teaching in colleges and universities. Firstly, the piano string vibration equation model is established to analyze the acoustic features such as pitch and overtones; secondly, MEMS inertial sensors and infrared detection rods are used to collect the playing gesture data, and the fusion of fixed posture is realized by the IU-EKF algorithm; then, the VGG- 16 deep learning model is used to extract the statistical features in time domain, spatial characteristics, finger coupling features and auxiliary features, and to realize the recognition of hand gestures; and finally, the Magic King is performed in different playing versions. The speed-strength visualization analysis is carried out for different playing versions. The results show that the average values of finger angle measurements are 147.92°, 136.03° and 117.15°, respectively, with a maximum error of only 2.73%; the maximum angular difference between the three paths of finger movement is only 2.6 degrees; the velocity calibration method effectively matches the finger sliding and rebounding velocities within the 95% confidence interval; and the predicted values of the music skill evaluation model are highly consistent with the actual values. The system proposed in this study can accurately identify piano playing gestures, provide objective quantitative indexes for piano technique training, and put forward optimization measures from three dimensions: fingering practice, technical difficulty attack, and experience learning, which is of great significance to promote the modernization of piano education in colleges and universities.
