Image Segmentation of Chinese Gong-che Notation Sheet Music Based on Clustering Analysis

Abstract

In this paper, the concept, development history, main technology and application scenarios of optical music recognition technology are briefly introduced. The research status of optical music score recognition of staff sheet music and Gong-che Notation sheet music is introduced, and then the classification methods and different characteristics of various clustering algorithms are analyzed, and some application scenarios of them are also discussed. This paper focuses on the image segmentation of Chinese Gong-che Notation sheet music by using hierarchical clustering algorithm, and extracts the useful information from the music's image. The experimental results show that different image features correspond to different useful information extraction accuracy, and the height and width features of the connected component can get the highest accuracy in the limited information extraction of Gong-che Notation sheet music.