Abstract Virtual reality panoramic technology transcends the constraints of time and space, offering audiences immersive visual and auditory experiences that diminish the costs associated with exploring historical and cultural landscapes and enhance communication effectiveness. This study employs fisheye cameras to capture panoramic images of such landscapes. The collected images undergo correction via spherical projection and reflective planar unfolding mapping. Subsequently, an enhanced method for optimal stitching line selection is applied to ensure seamless transitions between spliced images. The process is completed by integrating image interpolation and projection algorithms to virtually recreate historical landscapes. Analytical assessment of the model’s reproduction effects focuses on point error and accuracy. A practical application involving a real historical and cultural landscape is examined, revealing a maximum control point deviation from the optical scan of 0.02m and a minimum of 0.0014m, which negligibly affects the landscape’s virtual representation. Reproducing an actual landscape image yielded a maximum measurement error of 0.02m, a minimum of 0.01m, and an average of 0.0175m, affirming the capability for precise replication. This paper introduces innovative approaches and methodologies for leveraging virtual reality technology in the accurate reproduction of historical and cultural landscapes, illustrating its practical utility and effectiveness.