Improved Normalization Approach for Iris Image Classification Using SVM
With the rapid improvement of information technology, security and authentication of individuals has become a greater significance. Iris recognition is one of the best solutions in providing unique authentication for individuals based on their IRIS struct
- PDF / 309,029 Bytes
- 7 Pages / 439.37 x 666.142 pts Page_size
- 32 Downloads / 191 Views
Abstract With the rapid improvement of information technology, security and authentication of individuals has become a greater significance. Iris recognition is one of the best solutions in providing unique authentication for individuals based on their IRIS structure. Iris normalization meant to extract the iris region and represent it in spatial domain, Daughman’s rubber sheet model is so far a standard and efficient method of implementing this process. In this paper, a low complex, simpler and improved version of rubber sheet model is proposed. The main aim of this method is to minimize the complex computations that were involved in the conventional rubber sheet model and to provide an equivalent performing approach with very less computations. Classification performance is evaluated with CASIA and IIT Delhi IRIS databases using SVM classifier. Keywords IRIS
Normalization Rubber sheet model
1 Introduction In present scenario, keeping the data secured and authenticated is most difficult and very important to any organization or individual. Most of the researchers are focusing and developing different new applications on biometrics. Among all the biometrics iris recognition is playing vital and unique role. This works according to the visual features of a person like furrows, freckles, corona, and rings. Due to the high variation of randomness in the above-mentioned features, recognition of iris is considered as highly problematic approach [1].
M. Shaik (&) Department of ECE, JJT University, Jhunjhunu, Rajasthan, India e-mail: [email protected] M. Shaik Department of EEE, MJCET, Hyderabad, India © Springer Nature Singapore Pte Ltd. 2018 A. Kalam et al. (eds.), Advances in Electronics, Communication and Computing, Lecture Notes in Electrical Engineering 443, https://doi.org/10.1007/978-981-10-4765-7_15
139
140
M. Shaik
This organ is externally visible; it has unique quality of epigenetic pattern which stables throughout the adult life. This versatile character makes very attractive to use this as an identifying individual biometric application. By using different image processing techniques, the unique iris pattern can be extracted from a digitalized eye image, one can encode them into a biometric template and can be stored in database. A working model for automated iris recognition system with generic information formulated, which was a successful system [2]. Other most of the popular recognition models were also proposed [3–7]. The Daughman’s system was tested under several studies, the result rate always lead to zero failure rate. By dong experiments on millions of irises, Daughman’s system always gives perfect identification of individuals. By doing experiment on 520 iris images Wildes et al. who established the prototype system also gave flawless performance reports. By taking the database of 6000 eye images another researcher Lim et al. proposed a model by having recognition of 98.4%. The main objective of this paper is to present low complexity based IRIS normalization method to minimize the complex compu
Data Loading...
