Development of smart security system for building or laboratory entrance based on human’s brain (EEG) and voice signals

• Main Authors: Norizam, Sulaiman, Mahfuzah, Mustafa, M. S., Jadin, Fahmi, Samsuri, Amran, Abdul Hadi, Wen Xiang, B. L., M. S., Najib
• Format: Article
• Language: English
• Published: Universiti Teknikal Malaysia Melaka (UTeM) 2018
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://umpir.ump.edu.my/id/eprint/20589/1/Development%20of%20Smart%20Security%20System%20for%20Building%20or%20Laboratory%20Entrance.pdf; http://umpir.ump.edu.my/id/eprint/20589/; http://journal.utem.edu.my/index.php/jtec/article/view/3501/2449

The drastic increment in cyber-crimes and violent attacks involving our properties and lives made the world become much vigilant towards ill-intentioned peoples. Thus, it leads to the booming of smart security system industry which relies heavily on biometrics technology. However, due to certain circumstances, some users may find the existing biometrics technologies such as fingerprint, palm, iris and face recognition are unable to detect the necessary data precisely due to the physical injuries of the users. Furthermore, the fact that these biometrics technologies are easily retrieved from the user and be used as counterfeit to access to the security system undetected. Thus, in this research, in order to enhance the existing security system based on the biometric technologies, the combination of the human physiological signals such as brain and voice signals will be employed in order to unlock the magnetic door entrance to the laboratory, building or office. This research has utilized mobile Electroencephalogram (EEG) headset and voice recognizer to capture human’s brain and voice signals respectively. The extracted features from the captured signals then are analyzed, classified and translated to determine the device command for the microcontroller to control the door entrance’s locking system. The high rate of classification results of the selected features of EEG and voice signals at 96.7% and 99.3% respectively show that selected features can be translated to command parameters to control device.