Robust Interactive Method for Hand Gestures Recognition Using Machine Learning

The Hand Gestures Recognition (HGR) System can be employed to facilitate communication between humans and computers instead of using special input output devices. These devices may complicate with especially for people disabilities. gestures defined as a natural human-to-human method, which also used in human-computer interaction. Many researchers developed various techniques methods that aimed understand recognize specific hand by employing one or two machine learning algorithms reasonable accuracy. This work aims develop powerful gesture recognition model 100% rate. We proposed an ensemble classification combines the most classifiers obtain diversity improve majority voting method was aggregate accuracies produced each classifier get final result. Our trained self-constructed dataset containing 1600 images ten different gestures. canny's edge detector histogram oriented gradient great combination experimental results had shown robustness our model. Logistic Regression Support Vector Machine have achieved validated public datasets, findings proved outperformed other compared studies.