Principles of Laser Triangulation

Hermary scanners and many industrial 3D scanners work on the principle of laser triangulation. A laser beam is projected at a known angle onto a target to be measured; a camera at a known offset from the laser views the projected image.

What is triangulation?

Triangulation is a method for determining the distance between two points. It works by using two or more measurements taken from different positions. By measuring the angle between the two points and the distance between the two positions, it is possible to calculate the distance between them using trigonometry. Trigonometry is reliable because it is based on mathematical principles and formulas tested and proven over time. It is instrumental in applications such as surveying, navigation, and astronomy, where precise calculations are necessary for design and analysis. Many 3D scanners build on the principles of triangulation to examine the geometric relationship between the object and the scanner.

Why laser?

Laser is the acronym for “light amplification by stimulated emission of radiation.” In 1960, Theodore. H. Maiman invented the first ruby laser. It used a synthetic ruby crystal as the medium, which was excited by a high-energy flash lamp. The ruby laser emitted a deep red light with a wavelength of 694 nanometers, which was highly coherent and had a narrow beam. It was a major breakthrough and paved the way for future advancements in laser applications. The ruby laser was initially used for scientific research and was later developed for various industrial, medical, and military applications.

Modern commercial lasers are mostly made with semiconductor diodes, which allow for more precise and efficient laser production. When electricity flows through the diodes (p-n junction), it excites the atoms and causes them to emit photons, creating a laser light. Due to their highly concentrated and parallel (collimated) light beams that can reach long distances, lasers are perfect for achieving precision and accuracy on tiny spots. Additionally, red lasers’ exceptional visibility* is highly desirable for safety purposes. These two key features have made laser technology widely used in metrology and manufacturing.

*Note: Green lasers are more visible to the human eye than red lasers. Because green lasers fall in the middle of the visible light spectrum, the range that human eyes are most sensitive to. Despite this advantage, red lasers hold a bigger market share in commercial and industrial applications. As they are easier to build, making them economically feasible. They also require fewer components, contributing to their lightweight and compact size.

Common industrial scanning technologies using triangulation

The Scanners built on triangulation are non-contact, meaning they do not physically touch the target object to create a 3D representation. Non-contact scanning methods are fast and do not compromise the target object’s integrity. Three major scanning technologies capture an object’s 3D coordinates in space using triangulation:

  1. Laser triangulation
  2. Stereo vision
  3. Structured light

Each method is built differently, but all rely on the triangulation principle to calculate the depth or the distance to the object.

Laser triangulation

Laser triangulation scanners derive three-dimensional data using the principles of triangulation, with the laser being the illumination source. Laser triangulation scanning is an active sensing technology because it has an energy-emitting component. It can take measurements at a single point or across a scanning plane.

The scanner gathers data points in a reference plane established by a laser fan beam. All data points will be somewhere on this plane. If the laser strikes an object at A, this will be seen by the scanner’s image sensor at location A’. If the laser strikes an object at position B, it will be seen at B’. With the distance between the laser and the image sensor known, the scanner can calculate the distance to the object by examining where the sensor sees the reflected laser.

Principles of Laser Triangulation

Laser triangulation is extremely fast and can scan over 1,000 scans per second, delivering more than 500,000 data points (X, Y, Z). Laser triangulation finds applications in various industries, such as electronics, lumber production, food and beverage, and automotive, where its 3D data is used for inspection, feature identification, object detection, and process optimization. Since measurement requires movement of the object in relation to the scanner, laser triangulation is often used to monitor and help robotic tasks on fast-moving production lines. Laser triangulation may not be able to measure highly reflective surfaces. Still, due to its reliability and ease of integration, it remains a go-to method for collecting 3D data on conveyor belts and many other industrial settings.

A Typical 3D Scanner Setup

The setup below is a typical industrial setting. There are many 3D scanners that combine both the illumination source and the imaging sensor in one housing.

A typical laser triangulation setup.

Stereo vision

Stereo vision mimics human vision and uses a pair of cameras and overlapping fields of view to calculate an object’s distance from the scanner. Each camera takes a snapshot while an object is in view (a scene). The two images are compared to find common features in each scene. By knowing the distance between the cameras and the angle of their view, the object’s z value can be calculated using triangulation. Intel’s RealSense™ 3D cameras are built using the stereo vision principles.

Stereo vision is passive, meaning it does not emit energy to illuminate the object or scene. The cameras, therefore, can have a hard time in low light or when the object and the background have low contrast (e.g., white objects against a white wall). One of the ways to overcome this is by projecting a textured pattern onto the scene, called active stereo vision.

Stereo vision works well outdoors and is used in self-driving cars. Some applications include robotic navigation and laboratory environments, as stereo vision can provide full 3D range images out to a long distance. However, image processing for stereo vision is processor-intensive, and it may not be realistic in applications where high-speed or real-time results (e.g., a conveyor belt) is required.

A geometric representation of a typical passive stereo vision scanner.

Structured light

Structured light shines (projects) a known (pseudo-random) pattern on an object, and an offset camera captures the reflected light. The distance between the projector and the camera forms the baseline, which is typically factory-calibrated. The depth can be found using triangulation by examining (decoding) the deformed pattern. In reality, many scanners utilize two cameras and project multiple shots with different patterns during scanning. Binary and sinusoidal patterns are the most commonly used ones. Hermary’s DPS-4024HA coded-light scanner is also based on structured light technology.

Structured light-mounted robots are often utilized in manufacturing for inspection purposes. This scanning technology is also used in various handheld scanning devices beyond industrial applications. Microsoft Kinect v1 and iPhone’s Face ID are notable consumer products that use structured light technology. Handheld intraoral scanners rely on structured light for fast and accurate 3D dental representation.

Due to the potential interference of natural light, structured light systems are typically used indoors. Structured light is more effective in measuring objects at short distances as the projected pattern has a range restriction. Additionally, these systems may struggle with high-contrast or specular objects. It is worth noting that relative motion between the sensor and the scene can create distortions for the sensors. Though this issue can be improved with very fast projectors and cameras, structured light works best with stationary objects.

Relationship between resolution and scan distance

As the scan distance increases, the scanner resolution becomes coarser (more space between each data point). However, large objects (e.g., trees) can still be accurately measured by most triangulation scanners’ furthermost scan distance. Triangulation scanners are best when the target object is within the 3 m (10′) range, beyond which, Time-Of-Flight scanning should be considered. For best scanning results, please consult a professional System Integrator or an OEM company experienced in machine vision or contact us.

Learn more about 3D scanning: [Video] 3D Scanner Working Principles and How Point Cloud Works

3D machine vision, more broadly known as 3D scanning technologies, empowers inspection, feature identification, object detection, and process optimization by taking three-dimensional measurements of a target object. The most widely used 3D scanning technology is powered by the principle of laser triangulation.

3D data is often captured in coordinates, namely, X, Y, and Z. These coordinates collectively form point clouds that represent the 3D shape of the object being scanned. Depending on the scanner position and arrangement, further data manipulation may be required.

Watch more about 3D triangulation scanning: [Video] 3D Scanners Using Geometric Measurement Techniques

This video shows how laser scanners work seamlessly to capture a target object’s dimensions using the principle of triangulation. In modern manufacturing, laser triangulation is one of the most robust and reliable industrial scanning methods.