Image Feature Encoding Using Lownerization Tensor

Abstract

Nowadays, Technology is improving day by day. These days, vulnerability is the most important factor for any real-time application. There is a need to decrease the vulnerability of any application that is up to date. For many applications, till now we are using finger prints and faces as security. This paper advocates multimodal authentication using face and finger print.  The image analysis process is divided into three parts. Preprocessing of images, Extracting the features from the image and Classification and interpretation of features.  To save these extracting features, we need a technique so that these features cannot be vulnerable to anyone else. In this paper, Preprocessing can be done using Mean Square Error and Features can be extracted using Non-negative Matrix factorization. The proposed tensor technique called the Lownerization Tensor for encoding these extracted features