Study on the Construction of Immersive Learning Scene in the Construction of “Integration of Large, Medium and Small” Civics Course Based on Virtual Reality Technology

Abstract

The improvement of the integrated construction of ideological and political education promotes the equitable development of education. To enhance the realism and construction efficiency of immersive learning scenarios, this paper proposes an improved registration algorithm (RANSAC) based on the determinant point process (DPP) to optimize the sampling accuracy of image feature points. By establishing a collaborative education platform for primary, secondary, and higher education, a k-DPP sampling strategy is designed to enhance the uniformity of feature point spatial distribution, achieving high-precision learning scenario integration. Through comparative and control experiments, the effectiveness of the constructed immersive learning environment in assisting ideological and political education is validated. The study shows that the constructed technology achieves a maximum mIoU of 0.963 across 10 scenarios, with inference times ranging from 65 to 80 seconds, demonstrating good feature extraction performance and faster scenario construction speeds. The experimental class achieved an average satisfaction score of 9.52 for the immersive learning scenarios, with flow experience scores in all 8 dimensions exceeding those of the control class, and learning psychological states maintained at a high arousal level.