Contactless fingerprint recognition: A neural approach for perspective and rotation effects reduction
read more
Citations
Matching Contactless and Contact-Based Conventional Fingerprint Images for Biometrics Identification
Toward Unconstrained Fingerprint Recognition: A Fully Touchless 3-D System Based on Two Views on the Move
Touchless Fingerprint Biometrics: A Survey on 2D and 3D Technologies
Contactless Fingerprint Recognition Based on Global Minutia Topology and Loose Genetic Algorithm
Towards More Accurate Contactless Fingerprint Minutiae Extraction and Pose-Invariant Matching
References
Handbook of Fingerprint Recognition
Fingerprint image enhancement: algorithm and performance evaluation
Digital Image Processing 3rd Edition
Filterbank-based fingerprint matching
Fingerprint enhancement using STFT analysis
Related Papers (5)
Frequently Asked Questions (16)
Q2. What future works have the authors mentioned in the paper "Contactless fingerprint recognition: a neural approach for perspective and rotation effects reduction" ?
In future works, the authors should evaluate the performance of the proposed approach on datasets of contactless fingerprint images captured with a greater variability in the roll angle.
Q3. What are the steps of the biometric recognition process performed by contactless systems?
The biometric recognition process performed by contactless systems based on twodimensional samples can usually be divided into the sequent steps: acquisition; computation of a contact-equivalent fingerprint image; feature extraction and matching.
Q4. What is the reason why the minutiae on the borders of the ROI are removed?
The minutiae on the borders of the ROI are then removed because they are generated by false ridge-ends caused by the edges of the finger silhouette.
Q5. What are the parameters used for the enhancement algorithm?
The parameters used by the enhancement algorithm are ff = 0.2 and df = 20; the parameter hc defining the maximum considered height of fingerprint images is equal to 394 pixel (corresponding to 20 mm at a resolution of 500 ppi); the parameters adopted for the computation of the three-dimensional models used to simulate rotated fingerprint images are CH = 40 and CW = 2/5.
Q6. What is the proposed technique for the simulation of finger rotations?
The proposed technique for the simulation of finger rotations uses a synthetic three-dimensional finger model obtained from the ROI image RN , and then it computes a rigid transformation in the three-dimensional space.
Q7. What is the simplest technique used by twodimensional systems?
The simplest acquisition technique adopted by twodimensional systems consists in the use of a low-cost CCD camera in uncontrolled light conditions.
Q8. What is the generalization capability of neural networks?
The generalization capability of neural networks allows performing a robust estimation of the roll angle difference between two contactless acquisitions with a very limited need of computational resources with respect to traditional estimation techniques.
Q9. What is the common method used to perform biometric recognitions?
In order to perform biometric recognitions based on well-known methods in the literature, the computation of contact-equivalent images is usually performed.
Q10. What is the biometric technique for detecting fingerprints?
Most of the biometric technologies based on contactless fingerprint images perform the recognition task by using methods based on minutia features since they usually permit to obtain more accuracy with respect to algorithms based on global features.
Q11. What is the setup used to capture the contactless fingerprint images?
The setup used to capture the contactless fingerprint images is composed by a Sony XCD-SX90CR CCD color camera and a blue led with a light diffuser.
Q12. What is the significance of the proposed method for the detection of fingerprints?
The obtained results are shown in Fig. 6 and prove that the proposed method can increase the matching score between genuine individuals by effectively reducing problems related to different roll angles of contactless fingerprint samples.
Q13. How many nodes did the neural networks with a hidden layer achieve?
It is possible to observe that neural networks with a hidden layer composed by 40 nodes obtained the best accuracy on the considered dataset, with a total classification error equal to 1.65%.
Q14. What is the technique used to compute the ridge orientation map?
Similarly to the technique presented in [25], the method described in [7] applies Gabor filters tuned according to the local ridge frequency and orientation, but it computes the ridge orientation map by using an iterative regression algorithm designed to be more robust to the noise present in contactless fingerprint images.
Q15. What are the common techniques used to improve the visibility of the ridge pattern?
For this reason, most of the contactless recognition systems in the literature use illumination techniques to improve the visibility of the ridge pattern, like a point light source [8–10] or ring illuminators [19].
Q16. What are the systems that permit to obtain good quality fingerprint samples?
The systems appertaining to the first class permit to obtain good quality fingerprint samples, overcoming problems related to perspective distortions and finger rotations.