A roadway recorded by the vision stack, with vehicles marked and given a position in the scene.

Revolutionizing Vision. Reinventing ore mining, commerce, automotive and robotics.

The closest thing to a human eye for today’s machines. Not LiDAR.

SPI

Light and structure, in one stream.

Spatial Phase Imaging is a 3D vision format. The sensor records what a surface looks like and where it sits. FiguraOS reads that stream.

The sensor, seen through its aperture, with the imaging surface lit in color.

Sensor

  • Format

    Spatial Phase Imaging

    Light and structure in one stream.

  • Capture

    The sensor

    High-resolution visual and 3D data, live.

  • Software

    FiguraOS

    People, pose, and form on that stream.

  • Record

    Mesh and surface

    Pixels on target. Texture on the mesh.

FiguraOS

Software that reads shape.

Named for the Latin figura, shape. FiguraOS runs vision models on the live Spatial Phase Imaging stream: people, pose, and the form of what is in frame.

FiguraOS with a person marked for detection beside a pose skeleton and a position readout.

3D data

Mesh, pixels, surface.

A face shown as a high-resolution mesh, pixels on target, a triangle mesh, and a surface texture map.

Drag to rotate · Ctrl + scroll to zoom

Sample model for illustration. Not MAT Labs sensor output.

High-resolution mesh. Pixels on target. Triangle mesh. Surface texture map.

Markets

Where machines have to see.

MAT Labs supplies U.S. enterprises and U.S. government and defense programs.

A Komatsu haul truck driving up a road inside an open-pit mine.

Vision for the haul road.

Distance and motion for heavy equipment, in dust, glare, and low light.

Talk to us about Ore Mining

Ore Mining photo: Diego Delso, CC BY-SA 4.0, via Wikimedia Commons.

Company

μάτι

máti · Greek for eye

Named for the eye.

MAT comes from μάτι, máti, the Greek word for eye. The company is named for the work: the closest thing to a human eye for the machines of this era.

μάτι

Eye. The word the company is named for.

Figura

From the Latin figura, shape. FiguraOS is how the eye reads form.

Vision ex Machina

Vision, from the machine. A play on deus ex machina. Nothing is lowered in from outside. The machine produces the sight.

Capabilities

What the stream carries.

  1. 01

    People detection

    A person in the scene, marked on the live frame.

  2. 02

    Pose tracking

    A skeleton on the body, updated as the body moves.

  3. 03

    Face and body mesh

    High-resolution geometry of the person, not a flat outline.

  4. 04

    Pixels on target

    The image samples that land on the subject.

  5. 05

    Surface texture

    A texture map of the surface, aligned to the mesh.

  6. 06

    Live stream

    Capture and models on the same clock. The frame is not stored first and read later.

Specifications

The record.

What the stack produces. Numeric specifications are issued per program and are not listed here.
FieldRecord
FormatSpatial Phase Imaging
Sensor3D and visual, live
SoftwareFiguraOS
PeopleDetection and pose
MeshFace and body
SurfaceTexture map
PixelsOn target
TimingReal time
FiguresIssued per program

Contact

Request a demo.

Name the system and what it needs to see. Your request goes straight to our team.

Vision ex Machina

Vision, from the machine.