Robotic imitation of human hand using rgb color gradient mapping

Abstract

The state-of-the-art developments in robotics reveal a scenario wherein human motion imitation has been developed through advancements in features and performance of sensors. Most of such robots uses flex sensors to acquire the information about the orientation of the body being imitated. However, these are slow and are highly vulnerable to damages and physical changes. There are image processing algorithms that are focused on a robots learning through ANN and are not into imitation without memory. This paper presents an overview of research attempts to use a robotic assembly to imitate a real object using image processing. The paper describes results in force, tactile and visual sensing, sensor-based control, and the configuration of a test-based robot integrated system for imitation of a human hand, using vision through image processing using RGB scheme; and one-on-one mapping for the acquisition of the current position of a body and orientation of the various parts of the robot in same fashion. This research has resulted in the development of a prototype capable of imitation and its applications are in the areas of medicine, defense, safety, and explorations for both research and commercial applications.