Valid Kick Recognition in Smart Trunks based on Hidden Markov Model Using Doppler Radar

Smart Trunk Opener (STO) is a concept of opening and closing a vehicle's trunk or door without using keys, automatic and hands-free. STO system provides a flexible approach to the user by recognizing and classifying the hand movement or leg kick of the user and transmits the intention of the authorized user to open or close the trunk or door. However, such STO system using conventional approaches are typically sensitive to vibrations, unintentional people walking close to the trunk. In this paper, we propose to use 2D-Doppler spectrogram feature extraction method to classify valid kicks using 24-GHz Doppler radar while rejecting false alarms.