arXiv subjects

Iliya Shofman

Publications and source records attributed to Iliya Shofman.

3 recordsLinked to original sources

arXiv2026

Design of a Compact Monolithic Catadioptric Lens for CubeSat Hyperspectral Imaging Payloads

Monolithic catadioptric lenses have emerged as a novel class of compact long-focal-length lenses. Fabricated from a single fused silica substrate with distinct optical prescriptions across annular zones, these systems fold the optical path within the substrate, achieving long effective focal lengths at a fraction of the track length of refractive equivalents. The monolithic architecture and low coefficient of thermal expansion provide inherent athermal performance and preserve optical alignment under the thermal and vibrational loads of nanosatellite launch and orbital operation. Despite these advantages, commercial offerings of monolithic catadioptric lenses are limited and no comprehensive optical design literature review has been published for this lens class. We survey of monolithic catadioptric design variants, including Schmidt-Cassegrain-derived, annular folded, and gradient-index configurations. Building on this design space, we also present a novel monolithic catadioptric objective lens optimized for a CubeSat pushbroom hyperspectral Earth imaging payload. Pushbroom hyperspectral imagers require a precision entrance slit at the focal plane of the objective lens, and maintaining slit-to-focal-plane alignment under launch shock and thermal cycling is a significant optomechanical challenge. The proposed design places the focal plane coincident with the back surface of the substrate, enabling the slit to be lithographically etched directly onto that surface. This approach eliminates the slit-to-objective alignment degree of freedom, yielding an integrated, mechanically robust, and thermally stable objective-slit module for pushbroom imaging. We present the completed optical design of a near-diffraction-limited F/3 f=100mm VSWIR monolithic catadioptric lens, assess imaging performance, and discuss preliminary fabrication and tolerancing considerations.

physics.optics↗

arXiv2025

FINCH EYE: The Optical and Optomechanical Design of a GRISM-based SWIR Hyperspectral Imaging Payload for a 3U CubeSat

Crop residue is an important metric used for agricultural land-use monitoring and climate science research. Estimating crop residue coverage is essential to sustainable agricultural practices. The University of Toronto Aerospace Team is developing FINCH EYE, the optical payload for the upcoming FINCH 3U CubeSat, to measure crop residue cover. We conceived of a novel ultra-compact push-broom architecture with a volume phase-holographic grism dispersive element to keep the design compact and simplify the mechanical assembly. The FINCH EYE will image hyperspectral data from 900nm to 1700nm at 10nm spectral resolution, with a spatial resolution of 100m, and a SNR of 100. In this paper, we will describe the optical design of FINCH EYE, which consists of a commercial objective lens, an InGaAs camera, and a custom lens-grism-lens spectrograph. We will also describe the optomechanical housing, emphasizing design features that facilitate proper alignment during assembly.

physics.optics↗

arXiv2024

Beyond the Visible: Jointly Attending to Spectral and Spatial Dimensions with HSI-Diffusion for the FINCH Spacecraft

Satellite remote sensing missions have gained popularity over the past fifteen years due to their ability to cover large swaths of land at regular intervals, making them ideal for monitoring environmental trends. The FINCH mission, a 3U+ CubeSat equipped with a hyperspectral camera, aims to monitor crop residue cover in agricultural fields. Although hyperspectral imaging captures both spectral and spatial information, it is prone to various types of noise, including random noise, stripe noise, and dead pixels. Effective denoising of these images is crucial for downstream scientific tasks. Traditional methods, including hand-crafted techniques encoding strong priors, learned 2D image denoising methods applied across different hyperspectral bands, or diffusion generative models applied independently on bands, often struggle with varying noise strengths across spectral bands, leading to significant spectral distortion. This paper presents a novel approach to hyperspectral image denoising using latent diffusion models that integrate spatial and spectral information. We particularly do so by building a 3D diffusion model and presenting a 3-stage training approach on real and synthetically crafted datasets. The proposed method preserves image structure while reducing noise. Evaluations on both popular hyperspectral denoising datasets and synthetically crafted datasets for the FINCH mission demonstrate the effectiveness of this approach.

eess.IV↗