Structured Light

Structured light scanners are a cornerstone of high-precision non-contact measurement in manufacturing. They are used for:

• First Article Inspection (FAI): Verifying that a newly manufactured part matches its CAD design within micron-level tolerances.
• Dimensional Analysis: Measuring critical features like hole diameters, surface flatness, and gap widths on complex assemblies, such as automotive body panels or turbine blades.
• Reverse Engineering: Creating accurate 3D CAD models from physical prototypes by capturing dense surface geometry. Systems in this domain often use blue or white light patterns and achieve sub-10-micron accuracy, operating in controlled environments to ensure measurement integrity.

This technology powers key features in mass-market devices by enabling fast, accurate 3D facial recognition and gesture control. Common implementations include:

• Face ID Systems: Apple's TrueDepth camera projects over 30,000 infrared dots to create a precise depth map of a user's face for secure authentication.
• Augmented Reality (AR) Effects: Mapping facial geometry in real-time to apply realistic filters, masks, or virtual try-ons for accessories.
• Hand Tracking: Enabling controller-free interaction with devices by modeling the position and articulation of fingers. These systems prioritize compact form factors, low power consumption, and real-time processing using specialized application-specific integrated circuits (ASICs).

In healthcare, structured light provides non-invasive surface capture for diagnostics, surgical planning, and custom device fabrication. Key uses are:

• Orthotics and Prosthetics Design: Capturing the exact 3D contour of a limb to digitally design and mill a perfectly fitting socket, drastically improving patient comfort.
• Dental Scanning: Replacing physical impressions by intraorally scanning teeth to create digital models for crowns, bridges, and aligners.
• Wound Assessment: Quantifying the surface area and volume of chronic wounds (e.g., diabetic ulcers) over time to track healing progress objectively.
• Surgical Guidance: Providing real-time 3D surface topology during procedures to assist with navigation and incision planning.

For robots operating in dynamic environments, structured light sensors provide dense, pixel-aligned RGB-D data critical for:

• Bin Picking: Identifying and precisely localizing randomly oriented parts in a bin for a robotic arm to grasp, using the detailed depth map to avoid collisions.
• Simultaneous Localization and Mapping (SLAM): Building a 3D map of an unknown environment while tracking the robot's position within it, especially useful in GPS-denied areas like warehouses or indoors.
• Obstacle Detection and Avoidance: Creating a real-time depth field around a mobile robot or autonomous mobile robot (AMR) to navigate around people and objects safely. These applications require robust performance under varying ambient light and often fuse structured light data with other sensors like inertial measurement units (IMUs).

Museums and archaeologists use structured light to create highly detailed digital records of artifacts and historical sites without physical contact. This process enables:

• Digital Preservation: Creating archival-quality 3D models of fragile objects, such as ancient pottery or manuscripts, that can be studied virtually, reducing handling damage.
• Virtual Restoration: Simulating and planning the restoration of damaged sculptures or architectural elements by analyzing the digital model.
• Online Accessibility: Publishing interactive 3D models for public education and scholarly research. Scanners used here often have large field-of-view attachments and high-resolution cameras to capture fine surface details like tool marks or erosion patterns.

An active depth-sensing device that measures distance by calculating the round-trip time of a modulated light signal. Unlike structured light's pattern deformation, ToF directly measures the phase shift or time delay of reflected light.

• Mechanism: Emits a continuous wave of modulated light (often infrared) and uses a specialized sensor to measure the phase difference between emitted and reflected light.
• Applications: Mobile device face unlock, gesture recognition, and short-range obstacle detection for robots.
• Trade-offs: Generally faster frame rates than structured light but can suffer from multi-path interference and reduced accuracy at longer ranges.

The simultaneous capture of color (RGB) and per-pixel depth (D) information, typically from a single integrated sensor. Structured light and ToF are the two primary technologies that create RGB-D data.

• Output: A pixel-aligned pair of an RGB image and a depth map, where each pixel's value represents distance from the sensor plane.
• Key Use Cases: 3D scanning, robotic bin-picking, augmented reality occlusion, and human pose estimation.
• Hardware Examples: Microsoft Kinect (v1 used structured light, v2 used ToF), Intel RealSense D400 series.

A passive 3D reconstruction technique that computes depth by finding correspondences between pixels in two or more images taken from horizontally offset cameras, mimicking human binocular vision.

• Core Process: Epipolar geometry and disparity mapping. Depth is inversely proportional to the disparity (pixel shift) between matched points.
• Advantages: Passive (does not emit light), works well in sunlight, and provides dense depth at medium ranges.
• Challenges: Requires texture for matching; fails on uniform surfaces (aperture problem). Requires precise camera calibration.

Light Detection and Ranging is an active remote sensing method that measures distance by illuminating a target with pulsed laser light and measuring the reflected pulses with a sensor. It creates highly accurate, long-range 3D point clouds.

• Scanning Mechanism: Uses mechanical or solid-state systems to scan the laser beam across the field of view.
• Comparison to Structured Light: LiDAR is superior for outdoor, long-range applications (e.g., autonomous vehicles, topography). Structured light excels at short-range, high-precision indoor scanning of fine details.
• Output: A sparse or dense point cloud of the environment.

The science of making measurements and deriving 3D information from 2D photographs. It is a passive, software-based technique that relies on identifying common features across multiple overlapping images.

• Key Algorithms: Structure from Motion (SfM) to solve for camera poses and sparse geometry, followed by Multi-View Stereo (MVS) to generate dense reconstructions.
• Applications: Cultural heritage preservation, drone-based surveying, and visual effects.
• Trade-off vs. Active Sensing: Highly flexible and scalable but computationally intensive and requires good lighting and surface texture.

While often conflated, these are distinct concepts. Laser triangulation is a simpler, single-line variant that is foundational to structured light systems.

• Laser Triangulation: Projects a single laser line onto an object. A camera, offset at a known angle (baseline), views the deformed line. Depth is calculated using simple triangulation for each point along the line.
• Structured Light: Generalizes this concept by projecting a complex, coded 2D pattern (e.g., dots, stripes, grids). This allows a single camera snapshot to capture depth across the entire field of view, not just a single profile.
• Relationship: A laser line scanner is a 2D profiler; moving it across an object creates a 3D scan. Structured light captures a full 2.5D depth map instantly.