FIND PRETEND BIOMETRIC MISTREATMENT IMAGE QUALITY ASSESSMENTS FOR SPOOFING DETECTION

The Face, iris and fingerprint are most promising biometric authentication system that can be identify and analysis a person as their unique features that can be quickly extracted during the recognition process. To ensure the actual presence of a real legitimate trait in difference to a fake selfpretended synthetic or reconstructed sample is a important problem in biometric verification, which needs the development of new and efficient protection measures. Biometric systems are vulnerable to spoofing attack. A dependable and efficient countermeasure is needed in order to combat the epidemic growth in identity theft. The biometric detection and authentication deals with non-ideal scenarios such as blurred images, reflections and also faked by the other users. For this reason, image quality assessment approaches to implement fake detection method in multimodal biometric systems. Image quality assessment approach is used to construct the feature vectors that include quality parameters such as reflection, blur level, color diversity, error rate, noise rate, similarity values and so on. These features are stored as vector in database. Then implement Multi level Support Vector Machine classification algorithm to predict fake biometrics.