Privacy Protection Enhancement System In Multi-Model Biometric Using Multi-SVM Approach

  Gurpreet Kaur Bath,  Deepinder Kaur

The biometric systems are the systems which used the information of the person to identify the person. The process of biometric systems is reliable and effective approach to search out the identity of the person. Currently, the research work is focused to design the multimodal system by using the different classification approaches. But the performance of the previous conventional approaches was not suitable to achieve the better recognition rate. To expel the certain shortcomings, MSVM (Multi-class Support Vector Machine) which relied on the kernel mapping (KM).The planned approach MSVM-KM is utilized for the multi-modal system. The purpose behind the use of MSVM-KM is to enhance the accuracy with the recognition of the human face, fingerprints and iris recognition. Firstly, proposed approach performed the preprocessing to expel the noise and to enhance the quality of the recognition process. After this, the proposed approach collaborated with GWT (Gabor Wave Transformation) that dependent upon the feature extraction of face, iris, palm, fingerprint, etc specifically for the classification. At last, the proposed approach concatenated with MSVM to improve the recognition rate. The simulation outcome on the basis of the metrics like the computational time (CT), recognition rate (RR) and TPR (True Positive Rate) is given. The outcome demonstrated that the proposed approach is better in above mentioned terms to enhance the recognition rate and to expel the computation time as compared with the previous used approaches.

Keywords- SVM (Support Vector Machine), MSVM (Multiclass Support Vector Machine), GWT (Gabor Wave Transformation) and MSVM-KM (Multicalss Support Vector Machine- Kernel Mapping ).

Unique Identification Number - IJEDR1901006
Page Number(s) - 31-38
Pubished in - Volume 7 | Issue 1 | January 2019
DOI (Digital Object Identifier) -   
Publisher - IJEDR (ISSN - 2321-9939)

  Gurpreet Kaur Bath,  Deepinder Kaur,   "Privacy Protection Enhancement System In Multi-Model Biometric Using Multi-SVM Approach", International Journal of Engineering Development and Research (IJEDR), ISSN:2321-9939, Volume.7, Issue 1, pp.31-38, January 2019, Available at :http://www.ijedr.org/papers/IJEDR1901006.pdf