Schubert 3D camera ‘one step closer to long-term goal of optical 3D recognition’
The company now wants to adapt the technology for applications in other industries, claiming ‘all sectors in the manufacturing industry stand to benefit from 3D recognition technology’.
“With the use of the latest hardware and optimized, high-performance image processing software, we succeeded in coming a step closer to the long-term goal of optical 3D recognition”, said Dr Abdelmalek Nasraoui, head of Image Processing Development, Schubert.
He said the 3D camera, based on its 3D scanner technology, ‘not only delivers outstanding accuracy’ but increases the efficiency in pick & place applications and the degree of automation in secondary packaging.
The technology is based on a stereo approach, where two surface sensors, along with integrated LED illumination, record gray-scale images from two different angles (stereoscopy).
A coded light pattern is used to highlight structural surfaces of the 3D objects.
“The 3D camera can be used wherever a 3D product recognition is required. Especially in the food sector, one often also needs to record the product height, control the product according to the height profile etc, in addition to the length, width and area dimensions,” added Nasraoui.
“In the food and beverage sector, the "handle into the crate" (Bin Picking) simplifies product feeding (no product separation is necessary), which can only be realized with a 3D camera.”
Gerhard Schubert began developing image processing 30 years ago. Gerhard Schubert wanted his packaging robots to learn to see. Together with his research team for image processing, this was achieved over time.
Since 1996, Schubert has been using line sensors for the Vision System. The first unit was a transmitted light scanner. This was followed by the reflected light scanner and color reflected light scanner. These scanner types will also be used in the future. They operate reliably in continuous operation and are easy to commission
Schubert 3D Vision program
The 3D camera was developed as part of the Schubert 3D Vision program focusing on the detection of partially overlapping products or products to be removed from containers.
After each removal of a product by the robot, the 3D camera takes a new image for re-analysis and for determining the next product to be picked up.
Schubert’s 3D camera complements the company’s portfolio of image processing systems for pick & place applications.
After each robot intervention, i.e. pick-up, the work area is recorded using the 3D camera and the exact position of the next product to be picked up is calculated. Current position data enables the robot to accurately pick up the products even if they have shifted or are stacked.
The image data and the evaluation are visualized at the control terminal.
Nasraoui said during the development of the 3D camera, it was important the hardware and software components were as similar as possible to the Schubert 3D scanner.
Thanks to its user-friendliness and integration into the TLM (top Loading Machine) packaging system, the technology ‘reflects Schubert’s philosophy of “single-source” modular plant engineering’.