Teaching the methods of object detection by robot vision

Abstract

This paper presents two methods of object recognition by a robot vision which is being used to educate the students. This includes the basic manipulation of objects using the robotic arm. It describes the procedures needed for setting up the appropriate environment for object recognition and for each method gives the basic examples of practical usage of robot vision. The part of robot vision is accomplished by using a USB camera and software named the Image Inspector for the image analysis, and the ViewFlex is used to interconnect the output data from the Image Inspector and robotic arm. The robotic arm used in the described methods is SCORBOT ER-4U, which is a didactic manipulator. The first presented method uses the blob analysis where the size of the blob conditions object detection and the second uses the pattern matching. The students should be aware of a core problem, which is the relationship between the real three-dimensional space and taken image of an object in the reduced robot working area. The right approach and wide teaching of methods for the robot vision will surely increase a chance for rapid improvement of the existing technology.