Biometric access for a secure smart robot

Biometrics, an important research field emerging fast, has growing challenges related to the secured access and security of the systems protected. In this paper, the secured access to a smart robot is proposed using biometrics. The robot can be accessed by recognizing the user's face image, acting as a password, and also operates on the voice-commands given by the speaker. The face recognition technique uses `Principal Component Analysis' and `Eigenfaces'. The voice-commands are processed in real-time. The commands are processed in real-time using an on-line cloud server. The voice-commands converted to text form are sent via the Bluetooth network to a micro-controller on the robot, where these are interpreted and the smart robot performs the desired operation accordingly. The performance evaluation results of the face image recognition and the voice-commands signal processing are encouraging. Combining both the features makes the robot more robust. Few limitations of both the techniques have also been discussed with proposed solutions.