International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 04 Issue: 07 | July -2017
p-ISSN: 2395-0072
www.irjet.net
Efficient Altered Fingerprint Analysis and Rectification of Distorted Fingerprint Ms. Vidya Ramesh Patil1, Prof. (Dr). B. D. Phulpagar2 1Student,
Dept. of Computer Engineering, PES Modern College of Engineering Pune, Maharashtra, India Dept. of Computer Engineering, PES Modern College of Engineering Pune, Maharashtra, India ---------------------------------------------------------------------***--------------------------------------------------------------------2Professor,
Abstract - Identifying person accurately is important aspect
recognition result is termed as negative recognition system however, which is much more serious than positive recognition system, since malicious users may purposely degenerate fingerprint quality to preclude fingerprint system from finding the real identity. A number of factors like degradation of fingerprint image quality, including small finger area, cuts and abrasions on the finger, wet or dry finger, dirt on the finger or sensor, and skin distortion[6].
in many application areas like criminal cases. Face recognition system should be fault tolerant to handle such cases. Fingerprint recognition system has suffering through Positive and Negative classifications. Multi label classification can give best feature. In positive classification, where the physical access control systems and user should negotiate for selfidentification. The false case classification broadly talks about low quality of images in case of user identification may authenticate malicious user. Distortion detection can categorized in two-class classification problem which can be solved using the registered ridge orientation map and period map of a fingerprint is used as the feature vector and a Lib SVM classifier is trained to perform the classification task. Distorted fingerprint rectification (or equivalently distortion field estimation) is viewed as a regression problem, where the input is a distorted fingerprint and the output is the distortion field. For such problems Detection and Rectification of distorted fingerprint is must.
Low quality fingerprint recognition result is termed as negative recognition system however, which is much more serious than positive recognition system, since malicious users may purposely degenerate fingerprint quality to preclude fingerprint system from finding the real identity. A number of factors like degradation of fingerprint image quality, including small finger area, cuts and abrasions on the finger, wet or dry finger, dirt on the finger or sensor, and skin distortion. The aftermath of low quality fingerprints depends on the type of the fingerprint recognition system. Those fingerprints we call Altered Fingerprint [6].
Key Words: Ridge Pattern, Nearest Neighbour Regression, Orientation Field Map, Ridge Orientation map, PCA, Ridge Density.
2. Literature Review [2.1] Hardware based Distortion Detection [5] Fingerprint acquisition detects modified fingerprint so that modified fingerprint can be rejected. Researchers proposed to detect inappropriate force using specially designed hardware Bolle et al. [5] proposed to detect immoderate force and torque exerted by using a force sensor. After this they have received outcome that monitored fingerprint acquisition improves matching performance. Fujii [6] has suggested approach to detect distortion by detecting distortion using a transparent film attached to the sensor surface. Dorai et al. [5] have proposed to detect distortion by analyzing the speed in video of fingerprint. This method has following limitations –
1. INTRODUCTION In the last forty-year the fingerprint recognition technology has immersed but there is several challenging problems present in fingerprint technology and Major problems can be working with low the quality fingerprints. It has been identified using FVC2006 data-set that Fingerprint matching accuracy determined over same algorithm among various data-sets. The NIST [4] has conducted the evaluation and observed that there are many differences between matching the fingerprints accurately. Plain, Rolled, and Latent fingerprint matching technologies has many problems while recognizing the images. Basically there are two types of recognition systems i.e. positive recognition system and negative recognition system. In positive recognition system, physical access control system and user should cooperate and identify .In Negative recognition system; the fingerprint was not made by the person indicated. In positive recognition system, if the quality of image is not up to the mark then it seems to be fail for legitimate users also which in turn results into inconvenience. Low quality fingerprint
© 2017, IRJET
|
Impact Factor value: 5.181
i) Fingerprint sensors should have classic video capturing capability; ii) Fails for existing database images and iii) Could not check distorted fingerprints before pressing on the sensor.
[2.2] Distortion-Tolerant Matching [4] In this method, every pair of fingerprint is compared. For minutiae-based fingerprint matching method there are different strategies for Distortion such as
|
ISO 9001:2008 Certified Journal
| Page 2055