Plenoptic Condensation: A Novel Approach to Generalized Scene Reconstruction
2026-07-20 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors introduce a new method called Plenoptic Condensation (PCon) for reconstructing 3D scenes from images. Their approach starts by creating simple, low-detail representations of the scene and then progressively refines these into more detailed and accurate models. This allows for capturing fine details like sharp edges and smooth reflections better than existing methods. They tested PCon on real-world images and showed it can measure small damages on a car more precisely than current top methods. The flexible detail level in their models helps improve rendering and scene understanding.
Generalized Scene ReconstructionPlenoptic Condensation3D ReconstructionReality ModelsRepresentational PowerSharp EdgesReflective SurfacesBenchmarkingNeRORT-Splatting
Authors
Brevin Tilmon, Alex DeJournett, John Leffingwell, Scott Ackerson
Abstract
We present a novel Generalized Scene Reconstruction (GSR) approach called Plenoptic Condensation (PCon). PCon uses a multi-stage reconstruction pipeline, initially converting images into "soupy" scene elements with low (representational) power, then adaptively condensing the "soup" into "structured" elements of higher power capable of efficiently representing, for example, sharp edges and smooth reflective surfaces. PCon scene models called Reality Models (Relms) enable spatially varying representational power, which is essential for high-fidelity rendering, measurement, and scene understanding. We showcase several in-the-wild PCon reconstructions captured with consumer phone cameras and drones. In one case called "Damaged Fiat", PCon is benchmarked against two state-of-the-art (SOTA) GSR methods: NeRO and RT-Splatting. Referring to Figure 1 below, PCon reconstructs the car hood more than twice as accurately as the SOTA methods. But more importantly, the local damage profile error for PCon is 35 um (0.035 mm), whereas the two other SOTA methods are essentially unable to measure the damage at all. Our project website is available at https://quidient.github.io/pcon-2026.html.