A novel graphical password clustering method for fault tolerance mechanism

Pixel value access control (PV AC) was introduced to deliver a secure and simple graphical password method where it requires users to load their image as their password which referred as  PassPix. PV AC extract PassPix to  obtain three-octet 8-bits Red- Green-Blue (RGB) value where it used as password to authenticate a user and it is a case sensitive character. The pixel value must be match with the record stored in database or otherwise, the user is failed to authenticate. This is a challenge for PV AC where users which prefer to store PassPix on cloud would unintentionally alter and as well as the pixel value due to media compression and caused faulty pixel. Thus, this research was aimed to adapt the digital image clustering (DIC) technique to fix the issue where the faulty PassPix would be recognized having the same group as the original PassPix. However, most of DIC were mainly developed for content-based image retrieval (CBIR) which having totally opposite characteristics from PV AC which become as the problem statement of this research and address with three objectives to  solve.  The method employed in this research is a mixture of theoretical analysis and prototype experiment where the DIC techniques were nominate theoretically based on PVAC requirements. Experiments conducted in this research were meant to finalizing the DIC algorithm technically and to test the new PV AC model that developed throughout this research. The newly developed model would become a referencing model for DIC adaptation on security system since most of DIC research was focused on CBIR. Significantly, the PV AC become more flexible that allow user authentication process being execute cloudily which is crucially demanding for client-server access control especially during current pandemic outbreak situation.