The Research And Implementation Of Staff And Music Identification In Music Teaching

As two important research topic in the field of music, music recognition and Chinese numbered musical notation identification have an important significance on music notation format entry, staff maker software, music retrieval systems and music scoring system. With continuous research and innovation in both algorithms, the value of them is highly increasing.On Chinese numbered musical notation recognition, based on the traditional image processing methods, a method of identifying Chinese numbered musical notation recognition had been put forward. First, it used the standard Hough transform to detect the inclination of the notation.Then,it used a rotation matrix correction inclination of image processing, and then combined with the probability of finding the Hough transform and projection to the bar, and then extract the target contour with a contour detection method which is based on FreeMan chain code, and finally comprehensively analysed of the geometric characteristics and the horizontal and vertical projection feature of each symbol, and designed an algorithm to identify each music symbol, illustrated a specific thought of the algorithm and steps.At last,it experimentally proved the efficiency and accuracy of the algorithm.On music recognition, based on the classical fundamental frequency extraction methods, we proposed a peak based on the normal distribution of the harmonic peak fundamental frequency extraction method. First, stressed the importance of the fundamental frequency extraction in music recognition, and then proposed an improved short-term energy zero rate double threshold decision algorithm for preprocessing of the audio signal, and then analysed the traditional method and improved harmonic peaks confidence in the fundamental law of the merits of the extraction areas, further raised normality harmonic method, explained the algorithm principle, the algorithm processes. The theoretical analysis and experimental results show that the traditional harmonic peaks, improved confidence method compared to a normal harmonic accuracy than the former method greatly improved noise immunity than the latter greatly enhanced the time complexity only a slight increase, taking into account the advantages of both, while improving the shortcomings of both.