A Fingerprint Based Novel Cancelable Biometric Authentication Technique

In recent years, the usage of diverse biometric technologies for reliable automatic human recognition has grown at an exponential rate. There is rising concern that, with the widespread use of biometric technologies, people's privacy and anonymity may be in risk. In contrast to passwords and credit cards, biometrics are linked to a person forever and cannot be altered if they are stolen. Biometric patterns should be altered using non-invertible and revocable transformations to create cancelable biometric templates in order to avoid theft. A summary of latest advancement in different cancelable biometric systems for biometric template protection is provided. The shortcomings and benefits of current cancelable biometric systems are explored and point out promising research areas for development in this rapidly evolving domain. Also present a model of generating cancelable fingerprint template using Reed Solomon error corrections, which can be used to authenticate a distorted biometric image. This approach provides an accuracy of 95%.