Overview
Laser triangulation is a low-cost lidar design that can deliver high accuracy and cost-effective performance. It is widely used for indoor service robot navigation. This article describes the lidar core components and explains the operating principle of lidar based on laser triangulation.
Four Core Components of Lidar
Lidar systems typically consist of four core components: laser emitter, receiver, signal processing unit, and a rotation mechanism.
Laser emitter: The laser emitter generates the measurement beam. During operation it typically emits in pulses.
Receiver: Light emitted by the laser illuminates an object and is reflected or scattered. The reflected light is focused by an optical lens assembly onto the receiver.
Signal processing unit: This unit controls the laser emission and processes signals received from the receiver. It calculates target distance using the received information.
Rotation mechanism: The three components above form the measurement core. The rotation mechanism spins these components at a stable angular speed to scan a plane and produce a real-time planar map.
Principle of Laser Triangulation
Current lidar measurement methods mainly include pulse time-of-flight, coherent detection, and triangulation. Time-of-flight and coherent methods require more demanding hardware but can achieve higher accuracy, so they are often used in military and high-end applications. Laser triangulation is lower in cost while providing accuracy sufficient for many commercial and civil applications, so it is widely used.
Laser triangulation works by projecting a laser beam onto the target at a fixed incident angle. The laser is reflected or scattered from the target surface and an imaging lens focuses the reflected light onto a position sensor such as a CCD (charge-coupled device). When the target moves along the laser direction, the spot on the position sensor shifts. The displacement of that spot corresponds to the target displacement. Through algorithmic processing, the spot displacement is translated into the distance from the target to a reference baseline. Because the incident beam and the reflected beam form a triangle, the method uses geometric triangulation to compute distances, hence the name laser triangulation.
Depending on the angle between the incident beam and the surface normal of the target, laser triangulation can be classified as direct (normal incidence) or oblique.
1. Direct (Normal-Incidence) Laser Triangulation
When the laser beam is incident perpendicular to the target surface, i.e., the incident beam is collinear with the surface normal, the configuration is called direct or normal-incidence triangulation.
2. Oblique Laser Triangulation
When the laser beam forms an angle less than 90° with the surface normal, the incidence is oblique.
The laser emitted from the source strikes the target surface at a certain angle relative to the surface normal. The reflected or scattered light is gathered by a lens at point B and imaged onto a photosensitive unit for detection.
The angle between the incident ray AO and the baseline AB is α. AB denotes the distance between the laser emitter center and the CCD center. BF is the lens focal length f. D indicates the limiting image position on the photosensor when the target is at infinity. DE is the displacement of the image spot on the photosensor relative to that limiting position, denoted as x. With the optical geometry fixed, α, AB, and f are known parameters. From the similarity of triangles △ABO and △DEB, the side length relations are obtained:
Therefore, we can derive:
If one axis of the CCD position sensor is aligned parallel to the baseline AB (assume the y axis), the laser spot pixel coordinates obtained by the algorithm are (Px,Py). The projected displacement x can be computed as:
Here CellSize is the pixel size on the photosensor and DeviationValue is the correction between the projected distance from pixel coordinates and the actual projected distance x. When the target has a relative displacement along the baseline AB, x changes to x'. Using the relations above, the target displacement y is:
Summary
Both direct and oblique laser triangulation enable high-precision, non-contact distance measurement. The oblique configuration generally offers higher resolution than the direct (normal-incidence) configuration.
