ObjSplat: Geometry-Aware Gaussian Surfels for Active Object Reconstruction

TL;DR

ObjSplat tackles the challenge of autonomous, high-fidelity object reconstruction by unifying Gaussian surfels with geometry-aware viewpoint evaluation and a next-best-path planner. The approach jointly optimizes geometry and appearance, explicitly accounts for back-face visibility and occlusion in covisibility, and plans multi-step trajectories to maximize information gain while minimizing movement. Across simulations and real-world cultural artifacts, ObjSplat achieves complete, photorealistic models faster and more efficiently than state-of-the-art baselines, producing usable digital assets for physics-based tasks. The work demonstrates robust performance against varying geometry and texture, enabling practical, time-efficient robotic digitization with potential for broad XR and cultural heritage applications.

Abstract

Autonomous high-fidelity object reconstruction is fundamental for creating digital assets and bridging the simulation-to-reality gap in robotics. We present ObjSplat, an active reconstruction framework that leverages Gaussian surfels as a unified representation to progressively reconstruct unknown objects with both photorealistic appearance and accurate geometry. Addressing the limitations of conventional opacity or depth-based cues, we introduce a geometry-aware viewpoint evaluation pipeline that explicitly models back-face visibility and occlusion-aware multi-view covisibility, reliably identifying under-reconstructed regions even on geometrically complex objects. Furthermore, to overcome the limitations of greedy planning strategies, ObjSplat employs a next-best-path (NBP) planner that performs multi-step lookahead on a dynamically constructed spatial graph. By jointly optimizing information gain and movement cost, this planner generates globally efficient trajectories. Extensive experiments in simulation and on real-world cultural artifacts demonstrate that ObjSplat produces physically consistent models within minutes, achieving superior reconstruction fidelity and surface completeness while significantly reducing scan time and path length compared to state-of-the-art approaches. Project page: https://li-yuetao.github.io/ObjSplat-page/ .

Figures (12)

• Figure 1: ObjSplat autonomously plans viewpoints and progressively reconstructs an unknown object into a high-fidelity Gaussian model and water-tight mesh, enabling direct use in physics simulations.
• Figure 2: System Overview. ObjSplat progressively reconstructs unknown objects from RGB-D frames using Gaussian surfels as a unified representation. (Top left) Incoming frames are fused into the global model, where geometry–texture joint optimization, enforcing both photometric and geometric consistency. (Right) A geometry-aware view evaluation pipeline renders an uncertainty map by integrating occlusion-aware covisibility, surfel-wise confidence, and back-face detection to quantify surface quality and completeness. (Bottom left) Guided by this dense uncertainty, the next-best-path planner performs multi-step lookahead on a spatial topology, generating globally efficient trajectories that balance information gain and movement cost for active reconstruction.
• Figure 3: Identification of insufficiently reconstructed or uncovered regions. New Gaussian surfels are selectively added to regions exhibiting insufficient opacity, significant photometric discrepancy, geometric deviation, or back-facing surfaces.
• Figure 4: Co-visible views comparison. While conventional frustum-based metrics (Left) fail to account for self-occlusion, leading to false correspondences, our rendering-based check (Right) explicitly detects occlusion and back-facing surfaces, thus correctly identifying valid and invalid co-visible points, ensuring valid geometric constraints and confidence updating.
• Figure 5: Uncertainty qualification comparison on the Bunny Racer object. Our method enables fine-grained uncertainty evaluation with millisecond-level latency, effectively addressing complex geometric challenges such as self-occlusions and thin structures.