Piano music hybrid teaching mode utilizes multimedia technology to construct piano music teaching system, and adopts multimedia image denoising method based on wavelet transform to update the presentation effect of piano music digital resources. Combined with the characteristics of modernized piano music digital resources, a non-iterative clustering algorithm is proposed to innovate the sharing mechanism of piano music education resources. Personalized recommendation service for a multimedia piano music teaching system is realized by using a space vector characterization student interest model. With the help of relevant cases, learners are analyzed and the teaching content and objectives of piano music are set, and the regression analysis of blended learning mode and piano learning effect is carried out according to the overall learning situation of piano music. The average value of learning effect achieved by blended teaching of piano music is 3.962, and the regression coefficients of blended learning on the learning effect of cognition, skill, and affective attitude are 0.587, 0.572, 0.599 respectively, with the p-value being significant at the level of 0.001, which indicates that there is a positive correlation between blended learning and the learning effect of cognition, skill, and affective attitude. That is, the blended teaching method of piano music can improve the cognitive, skill, and emotional attitude levels of piano music learning.