Multi-modal compound biometric feature set security and person authentication using cancelable 2D color barcode pattern generation technique

This paper introduces the concept of two-dimensional (2D) color barcode, also known as color quick response (QR) pattern generation, and integration as an automatic method to produce the cancelable biometric template with improved recognition accuracy. It includes various methodologies for multi-modal based generation of biometric template, cipher conversion, diversity, irreversible property, etc. In this work, based on application of different attributes to four different biometric traits combining feature selection and fusion techniques, subsequently three templates are generated. However, in general, cancelable biometrics(CB) come with some systematic template distortion, which directly depends on input biometric characteristics to protect sensitive information. This will degrade the system performance when input deals with multiple biometric traits in the multi-biometric system. To address these issues, with the notion of color variant QR pattern analysis, dynamic constrained random key generation is introduced to generate CB templates. These templates can replace all other existing CB systems without compromising the quality metrics due to an independent transformation model for an authentication factor.