Monocular Online Reconstruction with Enhanced Detail Preservation

Fig. 1. Our pipeline processes a stream of monocular RGB images to reconstruct scenes with immediate feedback. Our method produces high-quality photorealistic maps with detailed reconstruction across multiple levels. The middle image illustrates our reconstructed mesh, and the right image showcases the rendered results of our reconstructed map, which captures high-quality details at both coarse and fine levels.

We propose an online 3D Gaussian-based dense mapping framework for photorealistic details reconstruction from a monocular image stream. Our approach addresses two key challenges in monocular online reconstruction: distributing Gaussians without relying on depth maps and ensuring both local and global consistency in the reconstructed maps. To achieve this, we introduce two key modules: the Hierarchical Gaussian Management Module for effective Gaussian distribution and the Global Consistency Optimization Module for maintaining alignment and coherence at all scales. In addition, we present the Multi-level Occupancy Hash Voxels (MOHV), a structure that regularizes Gaussians for capturing details across multiple levels of granularity. MOHV ensures accurate reconstruction of both fine and coarse

The project was completed during Songyin Wu's internship at Meta Reality Labs Research.