Cultural relics are important carriers of human history and culture, bearing rich historical information and cultural heritage. Accurately reconstructing and preserving cultural relics is an urgent issue to be addressed. Digital modeling effectively retains the visual features of cultural relics, making it significant in their reconstruction. Traditional visual reconstruction methods require multi-viewpoint images, which are time-consuming and inefficient. This paper employs the method of 3D Gaussian splatting for 3D reconstruction. However, the original method suffers from the presence of numerous floating artifacts in the reconstructed scene space, as well as artifacts in regions with insufficient initial point clouds, significantly degrading the reconstruction quality of cultural relics from different perspectives. In an effort to tackle this challenge, the present study introduces an innovative 3D Gaussian splatting technique that integrates depth information and diffusion models, Depth Diffusion Gaussian Splatting. Differing from the original approach, the proposed approach achieves significantly higher reconstruction quality while effectively reducing the occurrence of artifacts. It provides researchers with more accurate and reliable 3D models for the analysis of cultural relics, making a greater contribution to the field.