We’re happy to present what our research team has been working on over the last couple of months: our latest R&D efforts in markerless 3D object recognition for industrial applications, applying computer vision and depth-sensing technologies to enterprise environments.
How to achieve robust recognition of complex industrial equipment with depth-sensing cameras
Maintenance and training on industrial equipment is an important use for Augmented Reality. However, the objects themselves are often textureless and their real-world environments usually offer suboptimal conditions, such as inconsistent illumination.
These factors introduce significant challenges for reliable object recognition and tracking. Until now, most developers and solutions addressed these challenges by requiring the attachment of a marker.
We at CraftAR show you the next generation of 3D object detection in enterprise environment – getting rid of the markers.
We believe that reliable object recognition technology can make tasks like Manufacturing or Maintenance more efficient in the Aerospace, Automotive and Machine Industries, among others.
When combined with Augmented Reality, the recognition can highlight and superimpose additional information aligned with the real-world equipment. This helps companies and field workers reduce errors and make informed decisions in real-time.
Watch a demo video of our new markerless 3D object recognition tech
CraftAR has used several commercial depth-sensing cameras – including Intel® RealSense and Occipital® Structure Sensor – when approaching markerless 3D object recognition. We’ll share our related experiences at AWE, Santa Clara, June 1-2, and we plan to make our talk accessible to the public after the event.
Until then, here is our latest demo video. We ran our leading computer vision tech on a mobile device, with a Structure Sensor depth-sensing camera attached. The video presents how recognition technology can help to make a complex maintenance process more efficient.