Vision Online
Building a flexible picking system with 3D machine vision

03 Solution

The Zivid One+ 3D camera was chosen to precisely detect any target objects for Vision Online’s complete bin-picking solution. With the 2.3 MPixels 3D image sensor, the Zivid camera was able to capture RGB color, 2D, and 3D images simultaneously. It allowed customers to take images of any items, even challenging objects like shiny metal parts, with high accuracy. Thanks to the small form factor of the Zivid camera, customers were able to use it for both stationary- and robot mounting. The camera was small but covered a working distance from 30 cm to 3 meters, suitable for various bin-picking scenarios.

Vision Online used HALCON’s extensive standard libraries that are widely used in machine vision industries to simplify machine vision application development. The powerful HALCON libraries allowed users to calibrate the system with any TCP/IP robot speaker and define the searching zone as well as criteria on a 3D environment. Developers were able to easily create pick models from a CAD file by selecting multiple features of the surfaces and import their gripper model to define pick positions and collision avoidance rules.

04 Result

In 2019, Vision Online combined cutting-edge robotic technologies and introduced CoPicker, a comprehensive bin-picking robot cell. The CoPicker includes

• Zivid’s high-performance industrial 3D camera
• MVTec’s developer-friendly HALCON libraries
• Universal Robot’s flexible UR+ gripper

With CoPicker, automation companies can reduce error rates and improve picks per hour. The integrated solution helps developers use proven high-quality systems and meet the demanding requirements of industrial automation applications.