International Journal of Scientific & Engineering Research, Volume 7, Issue 12, December-2016 855 ISSN 2229-5518 IJSER © 2016 http://www.ijser.org Comparative Study of FAST, MSER, and Harris for Palmprint Verification System Satya Bhushan Verma, Saravanan Chandran Abstract—Palmprint based biometric systems are growing in security systems because of its higher accuracy and its working speed. The FAST, MSER, and HARRIS, are mainly used for image mosaic, pattern recognition, image retrieval, and many other applications in image processing. In this paper, comparative study of FAST, MSER, and HARRIS, for Palmprint verification is carried out. The feature of Palmprint is extracted using three different methods FAST, MSER, and HARRIS, and the Palmprint verification and comparison is performed. The HARRIS detects most number of points than others. The MSER detects lesser points and FAST detects the least number of points in palmprint images. But the FAST takes lesser time for matching and MSER takes more time for matching compared with other two methods. Index Terms—Biometrics, Palmprint, FAST, HARRIS, MSER, Pattern Recognition ——————————  —————————— 1 INTRODUCTION iometric is the science of calculating and analyzing physio- logical and behavioural data. Biometrics recognition sys- tems use behavioural and physiological data for validation [15]. The fingerprint, iris, face, palmprint, ear etc. are classified as physiological characteristics and signature, gait, keystroke etc. are classified as behavioural characteristics [13]. These be- havioural and physiological characteristics are unique and can successfully apply for biometric applications. The security systems based on password are widely used for information protection, but the passwords need to remem- ber. The password based system identify only who use this password not identifying the person. Security system also, cer- tain patterns such as alphabets, numbers and special characters and minimum length of 8 characters are insisted while creating new password and such passwords are difficult to remember. Biometric verification system is a method for identifying the template by matching with the previously stored template in the database. It measures a difference between the inputs to the previously stored templates from database and calculates some matching scores. Then it makes a decision on the basis of an assumed threshold value. Palmprint identification has emerged as one of the popular and promising biometric modalities for forensic and commercial applications. Palmprint features are considered promising in identify people [16]. Palm is an inner surface of a hand and a palm contains three types of lines that are flexion creases, sec- ondary creases, and ridges figure 1. The flexion creases are known as principal lines and secondary crease are called as wrinkles. Fig.1. The Line Patterns on the Palm Print Nowadays, there is high demand for touch-less biometrics due to various social and hygienic issues, Fig. 1 is the example of line pattern on the palm print. The biometric system based on palm print gives many benefits than the other biometric system, like it success fully works in lower dimension imaging, has a low cost, having stable structural feature, enhances fast feature extraction, higher accuracy rate, and great user adequacy. 2 RELATED WORKS A.H. M. Al-Melali et al. proposed a fast personal palmprint au- thentication based on 3-D multi wavelet transformation [1]. In that paper they used 3-D discrete multi wavelet transformation as feature extractor and a probabilistic artificial neural network as a classifier. They tested and evaluate their proposed method upon 240 palmprint images. They achieved 100% recognition rate. Chin-Chuan Han et al. [4] presented a paper of personal verifi- cation system which was based on the palmprint features. They extracted the ROI from palm image, and then extracted the B ———————————————— • Satya Bhushan Verma is currently pursuing PhD in Computer Science from National Institute of Technology, Durgapur, West Bengal, India. E- mail: satyabverma1@gmail.com • Saravanan Chandranhas completed Ph.D. from Department of Computer Applications, National Institute of Technology, Tiruchirappalli, in 2009. He is currently working as Assistant Professor at the National Institute of Technology, Durgapur, IndiaE-mail: dr.cs1973@gmail.com IJSER