With the advancement of Internet technology, people have diversified channels to obtain information. In the field of music, precise identification of different types and genres of music plays an important role in music creation and teaching. Therefore, this article proposes and studies a music recognition and computer-aided design (CAD) system based on artificial intelligence technology. The system uses an artificial intelligence analysis module to recognize music genres based on extracted features and deep learning algorithms, extract and reconstruct melodies, and automatically arrange and combine music elements. Research has shown that the CAD system designed for research has a higher accuracy in identifying music emotions than traditional music recognition techniques, with a maximum accuracy of 97.66%.