In the post-epidemic era, online classrooms and multimedia computers are increasingly connected. In order to efficiently conduct online teaching and answer questions after class, this paper intelligently collects and analyzes current multimedia computer teaching materials based on the latest audio-visual information fusion algorithm. First, the spatial frequency SF of the online classroom is selected as the input excitation of audio-visual information, and the direction information at the pixel is fixed as the link strength slice relationship of PCNN. Compared with traditional computer teaching, this method has many advantages. Multimedia electronic classrooms are very suitable for computer teaching, and the use of multimedia electronic classrooms can diversify the forms of computer teaching. Then, the source image is decomposed in a two-dimensional experience mode, and the obtained residual images are fused to complete the intelligent collection and analysis of the characteristics of teaching information; the function design is carried out under the framework of learning activities, and the functional upgrade of technical products requires the teaching and learning.