How Speckle-Free Lasers Improve 3D Inspections
Machine vision solutions that use laser displacement sensors are composed of three essential elements: a laser line generator, a camera for capturing the reflection of the laser line, and a computer for processing the image and extracting 3D data, and in some cases, 2D images as well. However, laser speckle has long plagued the accuracy of these systems and limited their ability to capture both detailed 3D point cloud models and 2D grayscale images.
2.5D height map (left) and 3D point cloud (right)
A laser displacement sensor works by tracking changes in a projected laser line as products pass through the plane of the laser. When the product passes through the plane, changes in the product’s surface are recorded as changes in the profile. Laser light is particularly well-suited to these 3D scanners because of its fundamental property of coherence. A laser beam will make the thinnest, least divergent line possible and be very bright. And when spread out into a thin sheet of light, like the fan-shaped beam of a laser profiler, the beam projects great distances with a thickness varying by only a few sheets of paper. This behavior comes at a cost, though.
The causes of speckle
The pure wavelength of the laser light allows it to interfere with itself and create random dark and bright spots in the image. This interference appears as grainy “speckle” when looking at a spot created by a laser. Speckle has been accepted as fundamental to laser illumination for 60 years and viewed as a problem in metrology for just as long.
In metrology, the impact of speckle is significant. The random dark and light patches cause a waviness in what should be a straight laser line on a flat part. This waviness, or uncertainty, contributes directly to a random height error that cannot be calibrated out. This is fundamental to laser optics. It is the fundamental limit of accuracy to any laser measurement system, which has afflicted in-line metrology systems for decades.