A novel authentication method for security systems using multidimensional characteristics of cost-effective flex sensors

摘要

Flex sensors are widespread in engineering and automation fields and their scope is increasing in biometrics and biomedical fields. The property exhibited by the flex sensors to change its resistance with the changes in physical flexing, can be used to construct an elementary authentication method based on the physiological characteristics of the human hand. Also, the changes in resistance can be associated with time dimension to implement a novel way of user verification based on the human hand dynamics. Inclusion of various time parameters such as rate of change of resistance or its further derivatives, along with space parameters (resistance variation) of flex sensors makes this technique reliable and unique. Furthermore, the computational needs are moderate as compared to other biometric systems or gesture recognition systems which generally use image processing and 2D/3D pattern recognition. This paper illustrates the development and testing performed for the mentioned intuitive approach which is centrally based on the uniqueness of the characteristics of the human hand. We developed a simple hand-glove mounted with flex sensors to realize the password authentication system with a modest cost of implementation. It can be used for user recognition in Human-Machine Interaction (HMI) or applied in the fields of Body Sensor Networks (BSN), biometrics, etc. This technique has potential applications as a wearable device with an advantage of being economical simultaneously.