3D Scanner Working Principles and How Point Cloud Works

Quality inspection, feature identification, object detection, and process optimization are some solutions empowered by precision 3D scanning. Understanding how 3D machine vision and point cloud data work can help system integrators and machine builders solve automation challenges that could not be addressed with 2D machine vision.

Video Transcript

Hi, I am Sofie.

Welcome to 3D machine vision for industry professionals. In this video, we will go over the working principles of 3D scanners using laser triangulation. We will also show you what 3D machine vision
data looks like and its benefits.

Now, a little about who Hermary is. Founded in 1990, Hermary is one of the leading
industrial 3D machine vision technology companies. We design and manufacture industrial 3D scanners that are used by some of the most brilliant system integrators and machine builders.

Through our 30-plus years of experience, we have participated in many projects that revolutionized automation in industries such as sawmilling, plywood, meat processing, industrial bakeries, pulp
& paper, steelmaking, warehousing and more.

Now, let’s take a look at three common types of 3D scanning technologies used in manufacturing.

The three main types of 3D scanning are interferometry, time of flight, and laser triangulation.

The Michelson interferometry is widely used in semiconductor manufacturing for quality assurance. The resolution of an interferometer is very high, but the range of measurement is very limited – only a few micrometers.

The most well-known use of time of flight is Lidar. In contrast to interferometry, the measurement
range can reach as far as a couple of kilometers. The resolution of Lidar is the lowest compared
to interferometry and laser triangulation. However, it is the key technology behind many surveying technologies.

At Hermary, we use laser triangulation to capture three-dimensional measurements by pairing a laser illumination source with a camera or an imager. In terms of range and resolution, laser triangulation
is the closest to what humans can measure using a tape measure or a micrometer. Laser triangulation scanners are calibrated to make measurements accurate to fractions of a millimeter.

Because of its robustness, laser triangulation has enabled many industries to capture critical 3D data that reduces waste, increases throughput and improves safety.

Now, I’ll hand over to my colleague, Steve. He is our 3D machine vision expert.

Hi, I am Steve.

I am excited to show you the working principles of 3D triangulation. Understanding this process can help you choose the right industrial 3D scanner for the job.

When scanning an object, the scanner is positioned so the laser beam and the camera – or the imager – are both aimed at the target object. As the target object passes through the scanner, the laser strikes the object at point A, a portion of the laser energy reflects off the target object and is imaged by the camera.

The scanner is factory calibrated during production with a built-in process to generate geometric XY data points of the target objects. Note that the shape of the laser can vary from collimated to fanned, capturing a single point or profiles of the target object. But the principle of laser triangulation remains
the same.

Depending on the application, the scanner configuration can range from one to many. Hermary’s GIG-4PT card is designed to connect up to 4 scanners and can be daisy-chained for applications requiring more scanners. A multi-scanner system is often used to overcome visual occlusion when scanning large items.

Now that we understand the working principles of 3D scanners, let’s look at the data they collect.

A point cloud is a list of 3-dimensional coordinates, namely, X, Y, and Z. Together, they form a digital representation of the target object’s surface. Point cloud data is a vector, meaning it can be manipulated, scaling, rotating, truncating or edited without compromising its integrity. This allows solution developers to truncate or filter the data to focus only on areas of interest.

Stitching is another powerful technique that combines multiple 3D scanner outputs into one common coordinate system. For example, a robot coordinate system, or a reference point of a machine. Vision-guided robots rely on this as they move relative to objects. Many system integrators and machine builders use 3D scanner data to create proprietary optimization and control software.

Point cloud’s deterministic nature ensures systems always output consistent and accurate results. 3D machine vision is most powerful when capturing free-formed items, large-scale objects, or objects with color variations. It is also often paired with a robotic arm to capture the object’s spatial data in relation to the environment.

Point cloud data has revolutionized the way many industries automate their processes. We are proud of the role Hermary technology plays in enabling levels of throughput and precision in manufacturing that go far beyond what is humanly possible.

For more information on Hermary’s 3D scanners or working with point cloud, visit our website or let us know in the comment section. Thank you for watching. We’ll see you next time.