In order to improve the effect of college English classroom teaching, this paper combines optical recognition to improve machine vision algorithms and uses computer optical vision technology to process teaching images. The system designed in this paper can be applied to philosophical teaching in college English classrooms. Moreover, the system can optimize teaching resources, manage teaching classrooms through the improved machine vision algorithm in this paper, and have a formative evaluation effect. In addition, taking into account the psychological activities of students in the classroom, this paper integrates the emotional recognition of college students in the construction of the system. Furthermore, it combines the actual teaching process to build a college English classroom teaching system based on constructivism. Finally, this paper designs an experiment to analyze the effect of the teaching model. From the research results, it can be known that the teaching system meets the demands of teaching.