Photonic imaging is the category of sensing systems that form spatial or spectro-spatial images of a scene by detecting and processing photons — as distinct from electronic (radar, ultrasound) or chemical (film) imaging. The category spans:
- 3D / ranging: LiDAR (direct ToF and FMCW), structured-light, and passive stereo — all relying on Time-of-Flight (iToF, dToF) measurement or geometric triangulation. SPAD arrays are the detector of choice for single-photon 3D sensing.
- Hyperspectral and multispectral: snapshot or scanning imagers that resolve a spectral dimension (near-IR through LWIR) across each pixel, enabling material identification and gas-column imaging. Mercury Cadmium Telluride FPAs dominate the thermal infrared bands; filter-wheel and Fabry–Pérot approaches serve near-IR.
- Thermal imaging: uncooled microbolometers and cooled HgCdTe arrays map scene temperature via MWIR/LWIR emission.
- Computational imaging: coded apertures, event cameras, and neural reconstruction pipelines that trade optics complexity for algorithm depth — relevant where SWaP constrains optic aperture.
The deep-tech investment angle is the integration wedge: moving from discrete optics + large focal-plane arrays to chip-scale photonic ICs with on-chip spectral filtering and pixel-level processing. Mid-Infrared Photonics integration drives cost reduction in spectroscopic imaging; Silicon Photonics underpins near-IR FMCW LiDAR.
Frontier
- Will chip-scale hyperspectral imagers reach cost parity with RGB cameras for consumer health wearables, or does pixel-count-versus-spectral-resolution tradeoff permanently stratify the market?
- Can computational imaging (coded aperture, event cameras, neural reconstruction) make low-cost sensors competitive with high-cost focal-plane arrays for industrial inspection?
- What integration level — monolithic vs flip-chip vs wafer-bond — best scales photon-counting SPAD imager arrays beyond 1 Mpixel at cost?