Smart camera guides pick-and-place robots
March 12, 2012
Omron Industrial Automation (www.Omron247.com) introduces a vision sensor developed specifically to meet the needs of pick-and-place applications. Highly compact, and with a processing speed of 5000 pieces per minute even with full 360 degree rotation, the new FQ-M vision sensor comes with Ethernet and EtherCAT embedded for ease of integration into any environment, and includes an incremental encoder for easy tracking and calibration. The FQ-M is easily configured for guidance of dynamic pick-and-place robots using Omron’s Sysmac Studio software, and is complemented by the palm-sized TouchFinder console for local monitoring and access to functions and settings.
The FQ-M is a smart camera that meets all the needs of demanding object recognition and tracking applications, combining a camera, image processing functionality and flexible communications options into a single compact package. The vision sensor ensures stable and robust shape-based detection even under changeable environmental conditions, and has been designed for the inspection of a wide range of objects.
The FQ-M vision sensor is the latest addition to Omron’s Sysmac automation platform, which tightly integrates core automation components with a common network and a single programming and configuration interface. It communicates with other devices either via EtherCAT – Omron’s network of choice for the Sysmac automation platform – or via standard Ethernet, or a combination of the two. For example, the FQ-M might be networked to an Omron NJ series machine controller for object tracking, and to the robot itself and the TouchFinder console over standard Ethernet.
Object detection is based on a newly developed contour-based search algorithm which ensures the highest reliability, regardless of changing lighting conditions, reflection and object inclination. The algorithm can even cope with overlapped or partially hidden objects. With this high performance algorithm and the high speed image processing capability, the FQ-M can detect up to 32 pieces at once with absolute stability, and 5000 pieces per minute. The incremental encoder interface simplifies initial calibration of the system, and enables on-the-fly tracking for tightly synchronized control. The FQ-M is able to output position coordinates and the corresponding encoder values, and is able to manage the object queue so that no object’s coordinates are duplicated.
Set-up of the FQ-M vision sensor is performed through the Vision Editor of the Sysmac Studio software, which provides intuitive, icon-driven configuration that makes it easy to achieve the optimum vision setting. The Vision Editor also provides comprehensive trending and logging functions. For easy integration into robot pick-and-place systems, the communication wizard makes it easy to configure any robot protocol either as a server or as a client without complex programming.