Search

The focus of this paper is on investigating the use of visual communication between two humanoid robots, in order to enhance the coordination of tasks between them. The problem continues to be an interesting and fruitful area of research from the days of using multiple robot manipulator arms for manufacturing as well as space robotics to current research in medical robotics. The approach here is to employ several off-the-shelf algorithms, software and hardware such as the NAO robot and support software, including Choregraphe, OpenCV to capture and process images, the SVM to classify objects in images, and the Python programming environment. Five robotic actions were studied and three modes. The experiments used one robot as the "viewer" and the second robot as the "subject" being analyzed. Results show that the visual communication system has an accuracy of 90% in correctly identifying the five movements. This research has shown an original solution, as a model that can enable robots to run in the complex service tasks consisting of multiple connected actions in a dynamic environment. This methodology can also let the intelligent operation of the robots serve in different scenes according to their actual requirements. This research focuses on enhancing the prototype robot vision function and development of additional value for consolidation manageable platform that increases service robots in the home environment of intelligent control capability. [ABSTRACT FROM AUTHOR]