Machine Control Using Hand Gesture Recognition
Gyancity Journal of Electronics and Computer Science · 2020
Human to machine collaboration is an important game changer in industries as well as in daily lives.Hand gesture recognition assists this interaction by automating the process.This collaboration is supported by proposing system which can recognize human hand gestures and perform tasks based on those gestures.This paper presents the system that uses hand gesture recognition to control machine.The machine executes programmed task according to gestures recognized.The hand gestures are recognized using vision-based approach.The defined gestures are static and carried out by bare hands.The recognition of the system involves hand segmentation, feature extraction, and classification.The hand is segmented using bounding box in webcam preview, and features are extracted by determining signature of hand and histogram of signature.The classification process uses supervised feed forward neural network system to train some samples of different gestures.Once gesture is recognized, the programmed tasks start to execute.The proposed system has found to have an average accuracy of 92.6% for the classification of gestures.