Robotic arm joint recognition in space using a neural network trained with pattern matching in LabWindows/CVI with Vision Development Module

Abstract

In this paper it will be presented a robotic arm joint recognition and control in space with cameras. The robotic arm is recognized in space and this way it can be controlled using the information from the cameras. The robotic arm's joints are recognized using a neural network which is trained with pattern matching in LabWindows/CVI with Vision Development Module. The pattern matching uses a previously learned picture pattern from the robotic arm's image and with image comparison can search for the specific pattern in the target image with the robotic arm. This way the robotic arm's joints can be located and controlled in space.