Ray Imaging

Le résumé fourni par la source

Chapter 4 explores ray-based imaging systems, beginning with geometric optics and progressing through pinhole and coded aperture imaging, projection tomography, and focal systems. It introduces visibility models for ray propagation and discusses how pinhole cameras form images without computation, though with limited resolution and sensitivity. Coded apertures improve throughput and enable 3D imaging by projecting structured shadows onto detectors, allowing for computational reconstruction using linear and neural methods. The chapter then delves into projection tomography, using Radon transforms and convolution backprojection to reconstruct object densities from ray attenuation data. It highlights the advantages of compressive sampling and neural inference in reducing exposure and improving image quality. Coded aperture tomography is extended to spectral and temporal domains, demonstrating snapshot compressive imaging (SCI) for high-dimensional data. Finally, focal imaging is introduced, detailing lens behavior, aberrations, and depth of field. The chapter emphasizes that optimal imaging design depends on signal characteristics, measurement constraints, and computational strategies.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.