Biometric Mistreatment Image Quality Assessments for Spoofing Detection


  • S. Sivaranjani
  • K. C. Ramya
  • S. Sheeba Rani


Face appearance, iris and finger impression are most promising biometric authentication system that can be identified and analysed a person’s unique features that can be immediately obtained from the recognition process. To confirm the real existence of an original authentic feature in difference to a fake or recreated model is an significant difficulty in biometric confirmation, which necessities the expansion of innovative and competent security methods. Biometric systems are susceptible to tricking attack. A trustworthy and well-organized counter measure is required to contest the epidemic growth in uniqueness holdup. The Biometric recognition and verification agreements with non-ideal circumstances such as distorted images, replications and also forged by others. For this motive, image quality valuation methods to instrument forged finding process in multimodal biometric systems. Image quality assessment approach is used to build the feature vectors that comprise quality parameters such as likeness, fuzziness level, color variety, error degree, noise degree, resemblance values and so on. These structures are stored as vectors in database. Then implement Multi level Support Vector Machine classification algorithm to predict forged biometrics.