A Visualization Method Of Music Based On Multiple Audio Features

All along, the music is an art form to express emotion, but listening to music in thetraditional sense can not satisfy people’s pursuit of the colorful world, Therefore musicvisualization arises at the historic moment. Also coincides with the computer technologyhas made great progress in image processing, virtual reality, digital signal processing.Music visualization technology in recent years like a duck to water, to get the attention ofresearchers and companies with real-world applications in entertainment, business,education and other fields. The music visualization organic combination of the auditorysystem and visual system, enhanced music experience feelings, to provide services for alltypes of people.With the increasing standard of living, people pay more and more attention to thespiritual and cultural life. When people enjoy music visualization, they want to see amore level varied visual effects, want to see the difference between different musicvisualization based on different types of music, want to better understand the gist of themusic. In order to solve these problems, we study the music visualization in term of themusic automatic classification, in term of multi-feature visualization method and in termof the main theme of the visual highlight.In the field of Music automatic classification methods, this paper put forward aautomatic classification based on the characteristics of music fractal dimension of music.In the first place one-dimensional fractal dimension characteristics of music has beencalculated and then build a music feature classification indicators to classify threedifferent musical styles, and with a different image elements corresponding to differentstyles of music.In the field which based on multi-feature visualization, this paper is based on style ofmusic has been classified then designs a comprehensive visualization methods based on aplurality of audio features. First for music multi-feature extraction, then these featurescomprehensive visual design, images to expression more music information. In the field of highlight the main theme of visualization methods, this paper design acomprehensive visualization methods based on a plurality of audio features. First formusic multi-feature extraction, then these features comprehensive visual design, imagesto expression more music information.Finally, the corresponding experimental data using various methods and effectsverify the feasibility and accuracy of each method.