The EEZTbot with computer vision
Abstract
Open Sourced projects are growing and gaining more and more attention. The idea to share knowledge so that it profits to everybody is also the spirit of this project. Computer vision is a task that is highly interesting for a lot of robotics projects however beginners can be lost and not be able to use this powerful tool. This project aims to use the open source robot manipulator EEZYbot MK1 of daGHIZmo with computer vision. A first part is made in order to simply manipulate it with a joystick. Then computer vision is explained with object detection. And finally, computer vision will be added to the robot so that the end effector can follow an object //WIP.
Installation
In order to manipulate the MK1 we used an arduino uno and a joystick as you can see on the following picture:
The servos are plugged on the PWN pins 5, 6, 9. And the joystick is plugged on the A0 and A1.
Conputer Vision
Steps are described on the github repository EEZYbotComputerVision so that the user can obtain the postion of an object in a square made with aruco markers as followed:
Jorand Gallou
Industrial and Aerospace Robotics Research Engineer
My research interests include robotics arm, machine learning and path planning.