Biometric identification system based on keyboard filtering

We have revised several authentication systems based on biometric technology to resume advantages and disadvantages. Because pure hardware biometric systems of user authentication have low rate on results over computational and economic cost, alternate biometric methods of low computational cost based on software development, are also being evaluated. We have developed a first prototype of a software system to elicitate sets of 20 password stroke samples, named attacks, with a population of 10 different users totalling 200 attacks. The results obtained demonstrate that users follow generally certain patterns when they are writing their password, and is possible to reinforce the user´s password authentication method by means of the analysis of user stroking patterns. In addition it is necessary to increase the population size and number of samples to establish standard and reliable rules. Finally, it is very difficult to find a general user pattern applied to every password