Big Blue Ceiling

Mixed-FOV Depth Sensing

Pairing an ultrawide depth sensor with a narrow one for detail where it matters.

Eddie OffermannHardware prototyping, enclosure, lens calibration, disparity-mapping algorithm, demonstration and report

Role
Hardware prototyping, enclosure, lens calibration, disparity-mapping algorithm, demonstration and report
Areas
Computer Vision, Hardware, Research, Optics
A black 3D-printed camera enclosure on a stand in front of a red curtain, with a small narrow lens and a larger fisheye lens side by side on its front.

Most depth setups pair matching sensors. This exploration broke with that to measure the effect of combining two depth sensors with different fields of view.

The question came from a hardware product that already had one ultrawide sensor. The study looked at what adding a narrower sensor would gain in detailed depth for the central region.

A two-column comparison. The ultrawide sensor was already in the hardware product and uses a large fisheye lens. The narrow sensor, added in the study, uses a small lens to give detailed depth in the central region.
The study measured what a narrow sensor adds to an existing ultrawide one.

We built the prototype hardware and its 3D-printed enclosure, which holds a small narrow lens and a large fisheye side by side, the lens calibration stack and a custom disparity-mapping algorithm, then demonstrated the results and wrote them up in a report.

A five-step flow: the two-sensor hardware prototype, a 3D-printed enclosure, a lens calibration stack, a custom disparity-mapping algorithm, and a demonstration and report.
The full stack built for the study, from hardware to report.

Part of Enhanced Presence.

  • A screen split in two: on the left a wide fisheye view of a man sitting in a room, on the right a tight, detailed view of the same man's face.
    The same scene from both sensors: the ultrawide view and the narrow, detailed center.
  • The black printed enclosure on a small tripod on a cork mat, its two lenses facing forward.
    The prototype on a desktop tripod.
  • The enclosure on its tripod seen from above at an angle, on a red patterned cloth, with its cable coiled beside it.
    The enclosure keeps both cameras fixed relative to each other for calibration.
  • The enclosure on a clamp and ball mount atop a light stand, in front of a red curtain.
    Mounted on a stand for testing.
  • The back of the black enclosure on its tripod, with a cable leading out of an open rear panel.
    The rear of the enclosure.
  • Close view inside the open enclosure, showing the camera boards' connectors and cables.
    Inside: the two camera boards and their cabling.
  • The enclosure seen from above and behind on a ball head and clamp, with its cables zip-tied to the stand, in front of a red curtain.
    The stand mount from behind.

  • Computational Optics

    Hyperspectral images from the ordinary camera in a phone.

    • Optics
    • Research
    • AI
    • Computer Vision
    • Hardware

  • HoloFlect

    A modular, upgradeable AR headset platform built for enterprise.

    • XR
    • Hardware
    • Computer Vision
    • AI
    • Optics
    • Research

  • ISO-5 Optical Cleanroom

    A cleanroom lab built to turn phone cameras into spectral instruments.

    • Optics
    • Hardware
    • Research
    • Computer Vision

  • Variable-Baseline Depth Sensing

    What happens to depth estimation when the cameras are very far apart?

    • Computer Vision
    • Hardware
    • Research
    • Optics