Pattern-Projection 3D Reconstruction for Industrial Shape Metrology
- Calibrated patterns (lines, grids, coded fringes) projected onto target encode local surface height via triangulation.
- Single-shot patterns for high-speed moving targets; multi-shot for higher accuracy on stationary inspection.
- Best fit for 3D inspection of stamped parts, solder paste, weld seams, food and pharma shape control and additive manufacturing.
- Projector–camera triangulation angle defines depth range; resolution scales with camera, pattern density and calibration.
- LED projectors reliable for industrial use; laser projectors brighter but introduce speckle that degrades resolution.
- Specular and absorbing surfaces require surface preparation or alternative techniques (photometric stereo, ToF).
Structured light projection extends machine vision into the third dimension by projecting calibrated patterns onto the inspected surface and analysing the deformation of the pattern in the camera image. The geometric relationship between the pattern projector, the camera and the inspected surface allows reconstruction of the surface height at every pixel, producing a three-dimensional point cloud that captures the full shape of the target. Structured light is the foundation of modern 3D machine vision and has displaced traditional contact measurement in many high-throughput applications.
Working Principle of Structured Light Projection
A structured light system consists of a pattern projector and a camera, mounted at a fixed angle relative to each other. The projector illuminates the target surface with a calibrated pattern, typically lines, grids, dot arrays or coded fringes. The camera, positioned at a known angle from the projector axis, captures the pattern as it appears deformed by the three-dimensional surface of the target. The deformation of the pattern at each pixel encodes the local height of the surface relative to a reference plane. Custom structured-light projectors are engineered within the Custom LED Illuminators portfolio.
The reconstruction algorithm processes the camera image to extract the local position of each pattern feature, computes the geometric correspondence between the projector and the camera, and triangulates the three-dimensional position of every point on the surface. The resulting point cloud, typically containing hundreds of thousands of points, provides a full three-dimensional representation of the target shape.
Pattern Types and Their Properties
Several pattern types are used in industrial structured light. Single-shot patterns (sinusoidal fringes, dot grids, random speckle) allow reconstruction from a single camera image and are preferred for high-speed inspection of moving targets. Multi-shot patterns (Gray-coded fringes, phase-shifted sinusoidal patterns) require sequential acquisition under different patterns and provide higher accuracy at the cost of longer cycle times. The choice between the two depends on the trade-off between speed and accuracy required by the application.
Typical Industrial Applications
Structured light is essential for three-dimensional inspection of stamped and moulded parts for shape, dimensional accuracy and warping; quality control of welding seams in automotive and structural assemblies; verification of solder paste deposition on printed circuit boards before component placement; measurement of profile and flatness on machined surfaces; inspection of food products for shape, volume and three-dimensional defects; quality control of pharmaceutical pills and capsules for shape consistency; verification of additively manufactured (3D-printed) components for shape accuracy; and any application requiring three-dimensional measurement at industrial throughput rates.
Selection Criteria and Design Considerations
The depth measurement range and the depth resolution are the primary specifications. The depth range is determined by the geometric configuration of the projector and the camera (specifically, the triangulation angle and the working distance). The depth resolution is determined by the camera resolution, the pattern density and the calibration accuracy.
The pattern projector must deliver high contrast and uniform intensity across the field of view, with the pattern features sharply focused at the working distance. LED-based projectors with engineered patterns provide good performance and high reliability for industrial applications, and compact directional projector heads are available from the LED Spot Illuminators family. Laser-based projectors deliver brighter patterns at the cost of speckle, which can degrade depth resolution.
Calibration and Algorithm Considerations
Structured light systems require careful calibration of the geometric relationship between the projector and the camera. Calibration typically uses a precision target imaged from multiple viewpoints, with the resulting parameters stored in the vision controller. Recalibration may be required periodically to compensate for mechanical drift, particularly in industrial environments subject to temperature variations and vibrations.
The image processing algorithm must be robust against surface texture, reflectance variations and ambient lighting. Modern algorithms include filtering of outliers, smoothing of the point cloud and detection of feature regions, all of which require parameter tuning for each specific application.
Integration and Limitations
Structured light systems integrate as standalone modules that include the projector, the camera and the dedicated electronics for synchronisation and image processing. The output of the system is typically a three-dimensional point cloud or a depth image, which is then analysed by application-specific software running on the vision controller.
The principal limitation of structured light is the dependence on surface reflectance. Highly reflective surfaces produce specular reflections that saturate the camera and prevent reliable pattern reconstruction. Highly absorbing surfaces produce weak signal that limits the depth resolution. For these surfaces, surface preparation (matt spray) or alternative techniques (such as photometric stereo or laser triangulation) may be required.
The second limitation is the geometric constraint of the triangulation: the projector and the camera must be at a non-zero angle to each other, which limits inspection in confined spaces or through narrow apertures. For such cases, alternative 3D techniques such as time-of-flight imaging may provide better integration despite their lower depth resolution.
RODER Vision Structured Light LED Projectors
RODER Vision engineers application-specific structured-light LED projector assemblies with calibrated pattern optics for industrial vision 3D reconstruction and shape metrology.
- Application-specific structured-light projector heads with calibrated patterns — Custom LED Illuminators
- Compact directional projector bases for fixed-pattern structured-light applications — LED Spot Illuminators
Synchronised pattern switching with camera trigger is essential for multi-shot acquisition — the RODER catalogue includes dedicated LED drivers and electronic controllers compatible with industrial machine vision controllers and PLCs.