ESAR: Event-Based Synthetic Aperture Reconstruction

2026-07-16

AI Summary

This work addresses the challenge of reconstructing large-scale structures of static scenes from monocular event cameras under near-nadir motion, where conventional approaches struggle. The authors propose a synthetic aperture inverse problem formulation based on a static log-radiance field. By enforcing geometric constraints that link dynamic latent views to the fixed scene, they integrate a linearized event accumulation model with geometric projection and temporal difference operators into a regularized inversion framework. This approach explicitly uncovers the synthetic aperture structure embedded in event data while circumventing direct estimation of the pixel-time volume. Experiments on both simulated and real-world Falcon Neuro datasets demonstrate that the method outperforms dynamic latent image-based and learning-based approaches, achieving more consistent large-scale spatial reconstructions and significantly suppressing fine-grained texture noise.

Abstract

Event cameras report asynchronous polarity events when changes in log--radiance exceed a fixed contrast threshold, producing signed temporal contrast measurements rather than conventional image frames. We formulate monocular event-based imaging as a synthetic-aperture inverse problem for a static ground-domain log--radiance field $θ\in \mathbb{R}^{N_g}$. Instead of reconstructing a latent pixel-time volume $v \in \mathbb{R}^{N_pN_t}$, we impose the geometric relation $v=Pθ$, where $P$ maps the fixed scene into motion-dependent latent views. Aggregating events over finite time intervals gives the linearized model \[ APθ= b+η, \] where $A$ is a temporal differencing operator, $b$ contains signed binned event counts, and $η$ represents measurement and modeling errors. This decomposition exposes a synthetic-aperture structure: under near-nadir motion, successive projections are approximately shifted views of a common scene, while the composite operator $AP$ remains ill-conditioned because it combines spatial averaging with temporal differencing. We therefore use regularized inversion to recover $θ$. Numerical experiments on simulated data and real near-nadir Falcon Neuro event data show that the proposed $θ$-based formulation recovers coherent large-scale spatial structure, relative to dynamic latent-image and learned event-reconstruction baselines, while suppressing fine-scale texture.

Problem

• event camera
• synthetic aperture
• log-radiance field
• inverse problem
• monocular imaging

Innovation

• event camera
• synthetic aperture
• inverse problem
• log-radiance field
• regularized inversion