Published Paper Details:

Abstract: Most of the people in this world are known to gestures, which is a means of communication amongst humans. They say actions speak louder than words, hence communicating using gestures is also a strong method. Robots are used for industrial and domestic purpose which are often controlled by the remote controllers. The employment of gestures as one of the way to manage people providing ease and will be more useful for the physically challenged. Gestures provide a special correlative manner to speech for expressing their objective. The main purpose of gesture recognition is that to associate a specific human gesture and convey information to the user with respect to individual gesture. Gesture recognition from hand motions or postures may be a lively area in gesture recognition research for Human Computer Interaction (HCI). We propose a gesture controlled wireless robot that performs the task of a spy robot serving security and monitoring purpose. The aim of gesture recognition is to spot a specific human gesture and convey information to the user regarding each gesture. Image processing is the foremost effective method accustomed process image signals. Training Image Classifier using Tensor flow Object Detection API (application programming interface) is followed where Convolution Neural Network (CNN) technique, coordinate value based algorithm is involved. Wireless interface is developed between controller and the robot that allows it to work in directions like front, back, right and left as per the commands received.