The impact of 3D machine vision technology

The impact of 3D machine vision technology

The development of 3D cameras and 3D image processing technology has opened up a wide range of new applications. They serve as the "eyes" of machines, providing a 3D spatial representation that enables automated decision-making based on the position, size and orientation of an object. For this reason, 3D imaging technology has become an integral part of many automated industrial processes, e.g. in quality control, assembly lines or bin picking.

In an AZo interview, Dr Martin Hennemann of IDS Imaging Development Systems GmbH discussed the application potential of 3D image processing, outlining the features, benefits and key developments of the Ensenso S 3D camera series.

Martin Hennemann

What influence has 3D machine vision had on industrial processes, particularly automation technology?

3D image processing has been able to establish itself as a mainstream automation technology. In combination with robotics, it is used worldwide for tasks such as automated picking of containers, pick-and-place, assembly operations or in logistics for depalletizing as part of intralogistics applications, such as handling and storage of goods or deliveries. One of the main advantages of 3D imaging technology is that it can further reduce human interaction in industrial processes. For example, robots are often used in industry to automate complex tasks or heavy lifting, which 3D vision makes more flexible.

In addition, there are more and more robots that are specifically designed to work with or otherwise assist human workers. This is a rapidly growing and advancing area of research. One main goal is to equip robots with enough "visual sense" so that they can exhibit intelligent behaviours and respond adaptively to their environment.

Where do the Ensenso 3D cameras slot into the existing machine vision field, and what advantages do they offer?

The Ensenso range of 3D cameras is comprised of active stereo vision 3D cameras and the recently added Ensenso S series, which utilizes a 3D technology based on structured light. They are flexible enough to adapt to a number of applications, seeing regular use in both industrial and research applications globally.

Could you offer more detail on the new Ensenso S model and some of its individual technological features?

The Ensenso S is one of the latest members of the Ensenso portfolio and signifies a new entry point to precision 3D imaging at a lower cost than its counterparts. Despite its lower price point, the 3D camera features industry-grade design and interface options and a modern 3D sensing technology that utilizes laser pattern triangulation, i.e., a variant of structured light. A laser dot pattern is projected onto the scene — for instance, a pallet with a stack of boxes on top of it.

Ensenso S then images that scene, recording the positions of these laser dots by capturing where on the boxes the light from the laser dots is scattered back into the system. Using the Ensenso software intelligence, this information is then processed via an artificial neural network that identifies the laser points individually, assigns each laser point an ID and then uses the principle of triangulation to obtain distance values for each of the laser points.

The camera itself features a compact, robust industrial housing. The camera housing is assembled to IP65/67 specifications, meaning it can also be used safely in harsh environments. The back of the device includes industrial connectors, such as an M12 GigE data port for data transmission and an M8 port for power, trigger or flash connectivity. A comprehensive selection of features is also offered.

What kind of data can the Ensenso S capture, and how can this be processed into appropriate forms?

The Ensenso S has the capacity to deliver high-quality and detailed 3D point clouds. The camera’s laser pattern triangulation works with up to 85,000 laser points, has a view field of 60° horizontal x 50° vertical and 500 to 3,000 mm working distance. The laser illumination permits different exposure times, which even makes objects in motion measurable. In terms of speed, 20 point clouds per second can be accomplished with consistently high resolution at full projector power.

The camera can offer high quality, low noise data to identify details up to around a centimeter in size. It also provides a low artifact level in contrast to a range of other technologies, such as low-cost stereo cameras and Time of Flight cameras. Via the flexible programming interfaces, calibrated point clouds and comprehensive intermediate data are readily available in line with the user’s preference of processing software for application integration.

Could you offer an outline of example usage scenarios of the Ensenso S 3D camera?

There are numerous ideal usage scenarios for the Ensenso S. When it comes to logistics automation examples, the camera is excellent in terms of handling pallets and tracking the movement of goods around and in between busy areas like storage units. When featuring in transport and movement applications, the Ensenso S has the potential to detect objects obstructing the way of transport vehicles. It is also an optimal choice for package handling. For instance, a 3D camera can be positioned over a conveyor belt with moving packages and deliver the 3D data for tracking and analyzing the packages.

As previously mentioned, Ensenso S has been optimized for use in combination with robots in depalletizing applications or even object recognition tasks that necessitate a specific range of objects to be checked, identified and controlled. The 3D camera is also ushering in new application areas, for instance, in horticulture where robots can monitor crops, tend to plants, survey fruits and automatically pick them at the right time when they are ripe.

Automated assistance, where robots collaborate with humans or even replace them for specific tasks, is also a rapidly developing field. For instance, it automatically selects and transports specific groceries from the store on demand. In scenarios as complex and challenging as this, the Ensenso S can make sure that the situations ‘make sense’ to these intelligent systems.

What are the main points and features that you would like our readers to keep in mind in relation to the Ensenso S 3D camera series?

The Ensenso S epitomizes a new entry point into precision 3D imaging for a broad spectrum of potential applications. This new, compact 3D camera is extremely versatile, cost-effective, and can integrate 3D vision into applications that have previously used this based on required budget. The laser triangulation algorithm deliver point clouds with accurate 3D geometry and a low artefact level, especially when compared to the results that other 3D technologies are able to acheive.

For integration, the Ensenso S comes with industry-grade housing and robust, comprehensive SDK features – all of which make this new camera an advanced and highly interesting product in the market.