Three-dimensional Gaussian scene reconstruction method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a three-dimensional Gaussian scene reconstruction method and device, electronic equipment and a storage medium, and the method comprises the steps: firstly, obtaining a plurality of three-dimensional Gaussian distributions of a target scene; then, obtaining a two-dimensional projection weight of each three-dimensional Gaussian distribution, and selecting a key Gaussian from the plurality of three-dimensional Gaussian distributions based on the two-dimensional projection weights; next, performing Gaussian splitting on the key Gaussian to obtain a plurality of split Gaussian distributions, and performing depth re-initialization based on the plurality of split Gaussian distributions and the plurality of three-dimensional Gaussian distributions to obtain a plurality of dense three-dimensional Gaussian distributions; and finally, selecting a plurality of reserved Gaussian distributions from the dense three-dimensional Gaussian distributions based on the two-dimensional projection weights of the dense three-dimensional Gaussian distributions, and performing scene reconstruction based on the reserved Gaussian distributions to obtain a reconstructed scene matched with the target scene. Therefore, the overall efficiency of three-dimensional scene reconstruction and the compactness and quality of the finally reconstructed scene are obviously improved. 本申请实施例提出的三维高斯的场景重建方法、装置、电子设备及存储介质，方法包括：首先，获取目标场景的多个三维高斯分布；然后，获取每个三维高斯分布的二维投影权重，并基于二维投影权重从多个三维高斯分布中选取关键高斯；接下来，对关键高斯进行高斯分裂，得到多个分裂高斯分布，并基于多个分裂高斯分布和多个三维高斯分布进行深度重初始化，得到多个稠密三维高斯分布；最后，基于多个稠密三维高斯分布的二维投影权重从多个稠密三维高斯分布中选取多个保留高斯分布，并基于多个保留高斯分布进行场景重建，得到与目标场景匹配的重建场景，从而显著提高了三维场景重建的整体效率和最终重建场景的紧凑性与质量。