Searcharxiv⌕ Search

arXiv · 2611.13663

Overcoming Dynamic Range Limitations in Computational Hyperspectral Imaging

Abstract

Computational hyperspectral imaging offers compact spectral sensing, yet imaging high-dynamic-range (HDR) scenes remains challenging. Accurate spectral reconstruction requires encoded measurements to remain within the calibrated linear-response range and above the noise floor. Due to unequal spectral responsivities across the encoding mask, individual channels reach their reliability limits at vastly different incident signal levels. Measurement reliability therefore depends on both exposure time and encoding channel, making it challenging for a single exposure to preserve spectral information in highlights and shadows simultaneously. Here, we demonstrate a physics-guided multi-exposure fusion framework that operates directly in the encoded measurement domain. An analytical reliability prior and learned spatially varying, channel-specific adjustments jointly determine adaptive fusion weights. Measurement-domain spectral-angle supervision further constrains relative channel responses for downstream spectral inversion. The set-wise formulation accommodates varying numbers and configurations of input exposures. We further implement the framework on a 9-channel broadband-encoded stochastic hyperspectral camera, incorporating full-field sensor-response characterization and channel-aware reliability gating. Simulated HDR benchmarks show lower reconstruction errors than representative fusion methods, while controlled spectral measurements and practical HDR scenes confirm improved spectral fidelity and extended usable dynamic range. Measurements using an optical-density target further establish a 16.0-dB increase in spectral reconstruction dynamic range relative to the best fixed single exposure.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Qianqian Yu, Jinhang Lin, Zhihai Yao, Shuo Wang, Chaofan Weng, Xiang Hao. 2026-11-10. Overcoming Dynamic Range Limitations in Computational Hyperspectral Imaging. https://searcharxiv.com/abs/2611.13663

Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite Searcharxiv by Indiciae: BibTeX · RIS.

KEEP EXPLORING

Related papers

Controlling Mode-Locking in Liquid Light

Mode-locking is the process by which the phases of many cavity modes organize collectively into a stationary state. In lasers, mode-locking originates from the interplay of gain, loss, dispersion and nonlinearities, most commonly resulting in the emission of short pulses. However, pulse formation is prevented when the gain recovery is fast enough to suppress amplitude fluctuations. Mode-locking then produces an extended state of nearly constant intensity, whose dynamics is governed by the phase. Yet, a general framework for designing the phase of such extended photonic states remains lacking. Here we introduce a hydrodynamic approach to designing extended photonic states, in which the phase dynamics are understood and controlled as a flow of liquid-like light. The intensity, clamped by fast gain, plays the role of a constant density, the instantaneous frequency that of a flow velocity, and the programmable radio-frequency modulation applied to the laser acts as a potential landscape that controls the flow. We derive analytical solutions connecting the applied modulation to the instantaneous frequency and hence to the spectrum emitted by the laser. We further provide experimental evidence for the liquid light shaping framework by measuring the frequency comb spectra emitted by a fast-gain semiconductor laser around 1.6 micron under the control of a programmed modulation, and comparing them to the ones predicted by the analytical solution. The hydrodynamic picture of mode-locking provides a predictive design principle for frequency combs, pointing toward a programmable synthesizer of light, and establishes fast-gain lasers as an accessible, electrically driven platform for nonequilibrium fluids of light.

physics.optics↗

Backward-scattering photoelectron holography of a polar molecule driven by a sub-cycle laser pulse

We investigate photoelectron holographic interference in the polar CO molecule driven by a sub-cycle laser pulse using time-dependent Schrödinger equation simulations and a classical-trajectory-based analysis. The sub-cycle temporal confinement of the driving field enables the forward-scattering holographic (FSH) and backward-scattering holographic (BSH) interference structures in the photoelectron momentum distributions (PMDs) to be clearly distinguished. The FSH fringes exhibit a strong dependence on molecular orientation, which is associated with the orientation-dependent transverse displacement of the ionizing electron wave packet (EWP). Furthermore, the heteronuclear nature of CO results in different FSH fringe shifts for ionization occurring predominantly from the C and O sides of the molecule. In contrast, the curved interference structures in the negative parallel-momentum region are identified as BSH interference arising from the coherent superposition of direct and backward-scattered EWPs. The BSH fringe positions show weak dependence on molecular orientation and remain clearly visible after orientation averaging. We further show that BSH visibility depends on the carrier-envelope phase (CEP) and that the fringes remain discernible after CEP averaging. Moreover, the BSH interference pattern exhibits a systematic dependence on the internuclear separation, demonstrating its sensitivity to molecular geometry. These results highlight the potential of sub-cycle pulse enabled backward-scattering photoelectron holography for probing molecular structure and demonstrate its robustness against molecularorientation and CEP averaging

physics.optics↗

Active Contrast-Dissipation All-Optical Encryption

Optical encryption is meant to place secrecy in physical structure, yet adversaries can readily construct imperfect physical keys to probe the encryption system via trial-and-error iterations. Such systems inevitably leak target plaintext information and preserve invertible residual structural features, rendering the entire encryption scheme vulnerable to adversarial inversion attacks. Here, we report an active contrast-dissipation all-optical encryption architecture, which partitions wavefront modulation into M wavelength-private encryption layers and N shared decoding layers. Security rests on multi-dimensional physical keys that must match simultaneously, including illumination wavelength, precise spatial alignment and layer stacking sequence. With every key matched and authorized readout is preserved: the composite plaintext is reconstructed at a full-detector Pearson correlation coefficient (PCC) of 0.997. We further impose active contrast dissipation as a core security constraint. Under physical perturbation it suppresses the foreground-background intensity contrast, so that the slightest key mismatch collapses the target features and dissipates the plaintext into randomized, diffuse speckle within chaotic background noise, leaving no residual semantic leakage to invert. Under nanoscale wavelength detuning, a 1 degree in-plane rotation, a single-pixel lateral displacement or layer-stacking tampering, the plaintext vanishes and the PCC approaches zero, whereas conventional diffractive encryptors retain identifiable contours. This work provide a feasible and reliable solution for high-security information encryption, with great application potential in communication security, information anti-counterfeiting, and confidential data storage.

physics.optics↗