Accurate hand gestures recognition through plastic barrier based on Doppler frequency using latent Dirichlet allocation classifier

Hand gesture recognition is vital for medical, industrial, military, and educational applications. Many parameters must be known its effects, the cross-section area of the moving hand part, motion angle, and its velocity. Also, the sensor which is used for hand detection is essential to achieve this requirement. Most studies in the field of hand detection have only focused on the camera sensors, while the drawback of that sensor is not working in all environments. Recently, there has been renewed interest in using radar as a detection sensor representing a solution to this problem. In the present work, radar has been used as a detector sensor and the latent Dirichlet allocation (LDA) as a classifier. The current results highlight the recognition rate gained by the radar has on; 98.1% without a barrier, while 88% with it.