The basis for correct gripping is hand-eye calibration, so that the robot grips at the position where the camera sees the object. In order for the detection to work reliably and safely, reliable detection of the objects is necessary.
The basis for this is the new detection based on deep learning. As soon as the objects have been recognised using deep learning, the gripping positions and angles are determined in the 3D image. The cloud of points then determines which of the objects is at the top and depalletising begins. The neural networks predefined in EyeVision allow this procedure for a large number of objects. If an object is not yet known, it can be taught in quickly using the integrated learning tool, so that these objects are also reliably recognised.
Bin Picking with EVT
Bin Picking is becoming increasingly important. The EyeVision software in combination with a robot guarantees that the components are safely removed from the container. This is done by the powerful Deep Learning Recogniser. In connection with a 3D camera, this enables the components to be safely removed, even in the case of bulk goods. It contains numerous pre-trained objects.
With the Deep Learning Recogniser, the hand-eye calibration and the correction of the absolute accuracy by the EyeVision software, even the smallest components can be gripped in the entire work area. In connection with a 3D camera, this enables the components to be safely removed, even in the case of bulk goods, and contains numerous pre-trained objects.
Further information on Robotvision with EyeVision image processing is available on the website.