Abstract: The computer graphics course, which is very practical, is a requirement for
teachers of the theory and practice of combining teaching of professional basic courses, How to
improve teaching quality is the current focus of higher education, and how to objectively and
exactly evaluate the teaching quality have positive and important meanings for improving the
teaching quality, Taking advantage of BP neural network theories, This paper establishes a
reasonable and effective model for evaluating the teaching quality of computer graphics course
which is based on working the index system out. After that, in order to find the existed problems
in our teaching process, which provide a direction of teaching innovation, the teaching quality of
computer graphics course in our university was evaluated by using this model.

Key words: BP neural network, computer graphics, teaching quality, evaluate