Structured Light

Projects a known pattern onto a scene and computes depth from how the pattern gets distorted.

Structured light is an active 3D-sensing method: a projector casts a known pattern — stripes, a coded pattern, or an infrared dot array — onto the scene, and a camera views it from a different angle. Surface bumps and dips distort the pattern, and knowing the relative position of the projector and camera lets the system triangulate depth at each point. It doesn’t depend on surface texture, so it works even on a blank white wall, and it’s accurate at close range. Well-known examples include the original 2010 Kinect, which projects a near-infrared dot pattern, and the iPhone’s Face ID, which projects over 30,000 infrared dots. Reflective or transparent surfaces make the pattern disappear, and strong sunlight washes out its contrast, so structured light is mostly used indoors at short-to-medium range. Active stereo also projects a speckle pattern, but computes depth through stereo matching between two cameras — that’s the key difference between the two methods.

ExampleThe original Kinect projects a near-infrared dot pattern across a living room; its infrared camera captures how the dots shift around people and furniture, and from that shift it computes per-pixel depth to generate a depth map for motion-controlled games.

Also called

Structured-Light Camera, Coded Structured Light, Structured-Light 3D Scanning

Related

Speckle Structured Light · Active Stereo · Time of Flight · Depth Camera · Triangulation · Transparent & Reflective Object Perception

Sources

Wikipedia: Structured-light 3D scanner
Wikipedia: Kinect（初代结构光与 Xbox One 版 ToF） (Chinese)