In this paper, according to the duration, the melodic element stress degree, the occurrence frequency of elements, and the "Melody contribution" are defined. Based on "Melody contribution" to a lot of musical melody abbreviated samples; the statistical analysis method is applied to establish the Hidden Markov Model (HMM) of musical melody. Thedefinition of "position chord", puts forward the improvement of the hidden Markov model. The dynamic program, the Viterbi search algorithm is used to find the chord matching of the given melody; the experimental results show that: The correct rate of the method application of the chord matching for the given musical melody reached 87.5%.
展开▼