In the current music education, the music teaching method is too simple, boring, relatively backward, unable to attract the attention of the students, and the students gradually lose their interest in music courses. In basic education, music, as a highly practical subject, must create a good classroom atmosphere, make the classroom active, and enable students to better experience music, enjoy music, and like music. Therefore, in order to change this situation, this paper, based on scientific computing visualization, studies the application of computer music technology in basic education music teaching, and proposes a related computing method based on scientific computing visualization-image segmentation method. This review compares the spectrum extraction accuracy between the improved algorithm and the two traditional algorithms and finds that the average accuracy of the improved algorithm is 90.88%. It can be seen that the method proposed in this review is more suitable for the study of computer music technology in basic education music teaching. Applied research and most students are more interested in music teaching combined with computer music technology, so learning music teaching combined with computer music technology is very necessary. The development of computer music technology is closely related to the development of modern information technology.