Artificial vision algorithm for object manipulation with a robotic arm in a semi-autonomous brain-computer interface

Abstract

We propose an artificial vision algorithm for a semi-autonomous brain-computer interface (BCI). The interface was designed in such a way that users are able to manipulate a robotic arm to pick up an object from a table and place it in one of two possible locations indicated as goal disks, and the manipulation is performed without any concern about continuous control of the final effector. The implemented algorithm is used to obtain, in real time, the positions of the object and the disks in reference to the robot frame. The main techniques used in the proposed algorithm were color segmentation and homography transformation. The implementation of the algorithm allows to obtain the positions of all the items in the table, and it successfully performs pick and place tasks, setting the items on different positions across the table.